JordanExistingDataMemo

Cover Sheet.ppt* Jordan Lake Nutrient Response Modeling Project: Existing Data Memorandum Prepared for
Triangle J Council of Governments
and the Jordan Lake Project Partners by Tetra Tech, Inc. August 2001 Summer Average Uncorrected Chlorophyll a Concentration (Station CPF055C) * * *
CPF055E
CPF055C
CPF0880A
CPF086C
jData
Chart1
SEG
SEAS$
DATE
STATION$
S
CPF050A
W
CPF055C
CPF055E
CPF055ESUR
CPF081A1C
CPF086C
CPF086F
CPF087B
CPF087B3
CPF087D
CPF08801A
CPF0880A
CPF0884A
TP
TN
NPRATIO
4.00
30146.00
0.11
10.70
97.27
4.00
30173.00
0.10
0.71
7.10
4.00
30209.00
0.05
0.61
12.20
4.00
30228.00
0.10
0.83
8.30
4.00
30271.00
0.11
0.95
8.64
4.00
30305.00
0.11
1.04
9.46
4.00
30329.00
0.10
1.14
11.40
4.00
30369.00
0.14
0.97
6.93
4.00
30404.00
0.08
1.03
12.88
4.00
30434.00
0.08
1.13
14.13
4.00
30447.00
0.13
0.94
7.23
4.00
30474.00
0.09
0.72
8.00
4.00
30504.00
0.12
0.62
5.17
4.00
30531.00
0.12
0.71
5.92
4.00
30566.00
0.06
0.43
7.17
4.00
30600.00
0.07
0.83
11.86
4.00
30629.00
0.13
1.32
10.15
4.00
30664.00
0.18
1.03
5.72
4.00
30711.00
0.09
1.12
12.44
4.00
30742.00
0.16
1.16
7.25
4.00
30768.00
0.18
0.92
5.11
4.00
30791.00
0.15
0.78
5.20
4.00
30819.00
0.17
1.25
7.35
4.00
30847.00
0.09
0.61
6.78
4.00
30874.00
0.14
0.82
5.86
4.00
30910.00
0.06
0.41
6.83
4.00
30943.00
0.08
0.78
9.75
4.00
30980.00
0.05
0.57
11.40
4.00
31000.00
0.11
1.03
9.36
4.00
31028.00
0.23
1.18
5.13
4.00
31089.00
0.13
1.02
7.85
4.00
31134.00
0.14
1.24
8.86
4.00
31155.00
0.11
0.81
7.36
4.00
31169.00
0.15
0.85
5.67
4.00
31211.00
0.21
1.03
4.91
4.00
31219.00
0.21
0.71
3.38
4.00
31222.00
0.16
0.61
3.81
4.00
31239.00
0.10
0.61
6.10
4.00
31273.00
0.09
0.61
6.78
4.00
31320.00
0.05
0.64
12.80
4.00
31344.00
0.07
0.92
13.14
4.00
31370.00
0.10
0.97
9.70
4.00
31391.00
0.08
1.11
13.88
4.00
31428.00
0.19
1.50
7.90
4.00
31460.00
0.27
1.30
4.82
4.00
31496.00
0.13
0.96
7.39
4.00
31517.00
0.10
0.59
5.90
4.00
31551.00
0.18
0.91
5.06
4.00
31580.00
0.24
0.90
3.75
4.00
31616.00
0.10
0.51
5.10
4.00
31631.00
0.10
0.61
6.10
4.00
31680.00
0.10
0.51
5.10
4.00
31706.00
0.11
1.11
10.09
4.00
31735.00
0.19
1.40
7.37
4.00
31762.00
0.35
1.60
4.57
4.00
31784.00
0.22
1.55
7.05
4.00
31834.00
0.19
1.18
6.21
4.00
31860.00
0.15
0.96
6.40
4.00
31876.00
0.17
0.93
5.47
4.00
31909.00
0.10
0.73
7.30
4.00
31931.00
0.20
0.61
3.05
4.00
32014.00
0.06
0.81
13.50
4.00
32044.00
0.14
0.78
5.57
4.00
32065.00
0.10
0.98
9.80
4.00
32254.00
0.16
1.05
6.56
4.00
32296.00
0.11
0.63
5.73
4.00
32364.00
0.07
0.61
8.71
4.00
32413.00
0.10
0.88
8.80
4.00
32440.00
0.11
1.10
10.00
4.00
32616.00
0.08
0.80
10.00
4.00
32672.00
0.14
0.91
6.50
4.00
32721.00
0.10
0.80
8.00
4.00
32765.00
0.06
0.51
8.50
4.00
32799.00
0.06
0.68
11.33
4.00
32980.00
0.10
0.91
9.10
4.00
33029.00
0.06
0.79
13.17
4.00
33087.00
0.06
0.51
8.50
4.00
33129.00
0.05
0.61
12.20
4.00
33168.00
0.13
1.08
8.31
4.00
33339.00
0.06
0.85
14.17
4.00
33457.00
0.03
0.51
17.00
4.00
33828.00
0.05
0.64
12.80
4.00
34219.00
0.04
0.59
14.75
4.00
34548.00
0.04
0.61
15.25
4.00
34900.00
0.09
0.61
6.78
4.00
35207.00
0.04
0.58
14.50
4.00
35219.00
0.04
0.67
16.75
4.00
35248.00
0.04
0.31
7.75
4.00
35299.00
0.04
0.71
17.75
4.00
35332.00
0.06
0.55
9.17
4.00
35571.00
0.07
0.45
6.43
4.00
35590.00
0.04
0.65
16.25
4.00
35612.00
0.04
0.52
13.00
4.00
35646.00
0.09
0.41
4.56
4.00
35691.00
0.06
0.41
6.83
4.00
35963.00
0.05
0.77
15.40
4.00
35983.00
0.05
0.41
8.20
4.00
36012.00
0.02
0.31
15.50
4.00
36320.00
0.03
0.45
15.00
4.00
36362.00
0.06
0.51
8.50
4.00
36376.00
0.07
0.71
10.14
Date
Total Phosphorus (mg/L)
Total Phosphorus (Station CPF055E)
Date
Total Nitrogen (mg/L)
Total Nitrogen (Station CPF055E)
Date
N:P Ratio
N:P Ratio (Station CPF055E)
4.00
30146.00
0.11
15.80
143.64
4.00
30173.00
0.17
0.99
5.82
4.00
30209.00
0.06
0.63
10.50
4.00
30228.00
0.10
0.73
7.30
4.00
30271.00
0.17
1.06
6.24
4.00
30305.00
0.26
1.09
4.19
4.00
30329.00
0.25
1.35
5.40
4.00
30369.00
0.15
0.92
6.13
4.00
30404.00
0.23
1.20
5.22
4.00
30434.00
0.15
0.72
4.80
4.00
30447.00
0.19
1.10
5.79
4.00
30474.00
0.17
1.00
5.88
4.00
30504.00
0.30
1.50
5.00
4.00
30531.00
0.25
0.86
3.44
4.00
30566.00
0.12
0.61
5.08
4.00
30600.00
0.12
1.07
8.92
4.00
30629.00
0.16
1.38
8.63
4.00
30664.00
0.23
1.09
4.74
4.00
30711.00
0.20
1.14
5.70
4.00
30742.00
0.22
1.05
4.77
4.00
30768.00
0.19
1.06
5.58
4.00
30791.00
0.21
0.93
4.43
4.00
30819.00
0.20
1.20
6.00
4.00
30847.00
0.14
0.53
3.79
4.00
30874.00
0.19
0.90
4.74
4.00
30910.00
0.11
0.41
3.73
4.00
30943.00
0.10
0.92
9.20
4.00
30980.00
0.31
0.85
2.74
4.00
31000.00
0.18
1.03
5.72
4.00
31028.00
0.25
1.26
5.04
4.00
31089.00
0.16
1.24
7.75
4.00
31134.00
0.17
1.18
6.94
4.00
31155.00
0.22
0.93
4.23
4.00
31169.00
0.22
0.94
4.27
4.00
31211.00
0.30
0.99
3.30
4.00
31219.00
0.29
0.81
2.79
4.00
31222.00
0.23
0.81
3.52
4.00
31239.00
0.30
0.93
3.10
4.00
31273.00
0.11
0.61
5.55
4.00
31320.00
0.09
0.78
8.67
4.00
31344.00
0.13
0.78
6.00
4.00
31370.00
0.26
1.22
4.69
4.00
31391.00
0.09
0.98
10.89
4.00
31428.00
0.42
2.00
4.76
4.00
31460.00
0.27
1.11
4.11
4.00
31496.00
0.17
1.03
6.06
4.00
31517.00
0.14
0.71
5.07
4.00
31551.00
0.23
0.42
1.83
4.00
31580.00
0.33
0.93
2.82
4.00
31616.00
0.35
0.81
2.31
4.00
31631.00
0.28
0.61
2.18
4.00
31680.00
0.11
0.51
4.64
4.00
31706.00
0.14
0.94
6.71
4.00
31735.00
0.20
1.38
6.90
4.00
31762.00
0.43
1.60
3.72
4.00
31784.00
0.29
1.50
5.17
4.00
31834.00
0.16
0.97
6.06
4.00
31860.00
0.23
1.13
4.91
4.00
31876.00
0.19
0.90
4.74
4.00
31909.00
0.20
0.84
4.20
4.00
31931.00
0.19
0.61
3.21
4.00
32014.00
0.14
0.85
6.07
4.00
32044.00
0.22
0.96
4.36
4.00
32065.00
0.16
1.12
7.00
4.00
32254.00
0.23
1.19
5.17
4.00
32296.00
0.18
0.93
5.17
4.00
32364.00
0.24
0.71
2.96
4.00
32413.00
0.20
1.23
6.15
4.00
32440.00
0.12
1.04
8.67
4.00
32616.00
0.14
1.19
8.50
4.00
32672.00
0.16
1.03
6.44
4.00
32721.00
0.14
0.82
5.86
4.00
32765.00
0.11
0.61
5.55
4.00
32799.00
0.07
0.72
10.29
4.00
32980.00
0.16
0.99
6.19
4.00
33029.00
0.08
0.91
11.38
4.00
33087.00
0.08
0.86
10.75
4.00
33129.00
0.09
0.91
10.11
4.00
33168.00
0.14
0.98
7.00
4.00
33339.00
0.08
1.02
12.75
4.00
33457.00
0.06
0.71
11.83
4.00
33828.00
0.06
0.65
10.83
4.00
34219.00
0.07
0.70
10.00
4.00
34548.00
0.16
0.79
4.94
4.00
34900.00
0.11
0.63
5.73
4.00
35207.00
0.07
0.65
9.29
4.00
35219.00
0.07
0.94
13.43
4.00
35248.00
0.05
0.51
10.20
4.00
35299.00
0.05
0.61
12.20
4.00
35332.00
0.07
0.60
8.57
4.00
35571.00
0.10
1.04
10.40
4.00
35590.00
0.06
0.85
14.17
4.00
35612.00
0.06
0.88
14.67
4.00
35646.00
0.11
0.61
5.55
4.00
35691.00
0.06
0.21
3.50
4.00
35963.00
0.07
0.85
12.14
4.00
35983.00
0.07
0.42
6.00
4.00
36012.00
0.05
0.52
10.40
4.00
36320.00
0.07
0.55
7.86
4.00
36362.00
0.17
0.61
3.59
Date
Total Phosphorus (mg/L)
Total Phosphorus (Station CPF055C)
Date
Total Nitrogen (mg/L)
Total Nitrogen (Station CPF055C)
Date
N:P Ratio
N:P Ratio (Station CPF055C)
3.00
30146.00
0.05
7.40
148.00
3.00
30173.00
0.04
0.51
12.75
3.00
30209.00
0.03
0.61
20.33
3.00
30228.00
0.09
0.83
9.22
3.00
30271.00
0.06
0.74
12.33
3.00
30305.00
0.08
1.30
16.25
3.00
30329.00
0.05
0.88
17.60
3.00
30369.00
0.06
0.82
13.67
3.00
30404.00
0.06
1.01
16.83
3.00
30434.00
0.07
0.80
11.43
3.00
30447.00
0.05
0.72
14.40
3.00
30474.00
0.04
0.44
11.00
3.00
30504.00
0.04
0.41
10.25
3.00
30531.00
0.06
0.41
6.83
3.00
30566.00
0.04
0.41
10.25
3.00
30600.00
0.06
0.71
11.83
3.00
30629.00
0.06
0.86
14.33
3.00
30664.00
0.11
1.14
10.36
3.00
30711.00
0.07
0.98
14.00
3.00
30742.00
0.11
0.94
8.55
3.00
30768.00
0.08
0.81
10.13
3.00
30791.00
0.08
0.71
8.88
3.00
30819.00
0.05
0.65
13.00
3.00
30848.00
0.06
0.41
6.83
3.00
30874.00
0.12
0.41
3.42
3.00
30910.00
0.05
0.51
10.20
3.00
30943.00
0.05
0.61
12.20
3.00
30980.00
0.03
0.48
16.00
3.00
31000.00
0.04
0.61
15.25
3.00
31028.00
0.07
0.81
11.57
3.00
31089.00
0.08
0.82
10.25
3.00
31134.00
0.06
0.81
13.50
3.00
31155.00
0.06
0.68
11.33
3.00
31169.00
0.05
0.43
8.60
3.00
31211.00
0.04
0.51
12.75
3.00
31222.00
0.04
0.65
16.25
3.00
31239.00
0.04
0.51
12.75
3.00
31273.00
0.04
0.61
15.25
3.00
31320.00
0.04
0.60
15.00
3.00
31344.00
0.05
0.65
13.00
3.00
31370.00
0.08
0.89
11.13
3.00
31391.00
0.06
0.99
16.50
3.00
31428.00
0.05
1.01
20.20
3.00
31460.00
0.05
0.93
18.60
3.00
31496.00
0.06
0.82
13.67
3.00
31517.00
0.05
0.49
9.80
3.00
31551.00
0.05
0.62
12.40
3.00
31580.00
0.03
0.51
17.00
3.00
31616.00
0.04
0.41
10.25
3.00
31631.00
0.04
0.61
15.25
3.00
31680.00
0.03
0.71
23.67
3.00
31706.00
0.06
0.94
15.67
3.00
31735.00
0.03
1.08
36.00
3.00
31762.00
0.12
1.43
11.92
3.00
31784.00
0.19
1.51
7.95
3.00
31834.00
0.09
1.15
12.78
3.00
31860.00
0.11
0.74
6.73
3.00
31876.00
0.06
0.74
12.33
3.00
31909.00
0.04
0.52
13.00
3.00
31931.00
0.05
0.41
8.20
3.00
32014.00
0.03
0.41
13.67
3.00
32044.00
0.05
0.81
16.20
3.00
32065.00
0.08
0.83
10.38
3.00
32254.00
0.08
0.89
11.13
3.00
32296.00
0.04
0.51
12.75
3.00
32364.00
0.03
0.51
17.00
3.00
32413.00
0.05
0.61
12.20
3.00
32440.00
0.07
0.63
9.00
3.00
32616.00
0.06
0.67
11.17
3.00
32672.00
0.03
0.41
13.67
3.00
32721.00
0.04
0.41
10.25
3.00
32765.00
0.07
0.61
8.71
3.00
32799.00
0.08
0.58
7.25
3.00
32980.00
0.06
0.68
11.33
3.00
33029.00
0.04
0.63
15.75
3.00
33087.00
0.03
0.41
13.67
3.00
33129.00
0.03
0.51
17.00
3.00
33168.00
0.09
1.11
12.33
3.00
33339.00
0.04
0.61
15.25
3.00
33457.00
0.02
0.41
20.50
3.00
33828.00
0.02
0.53
26.50
3.00
34219.00
0.03
0.37
12.33
3.00
34548.00
0.02
0.61
30.50
3.00
34900.00
0.05
0.51
10.20
3.00
35207.00
0.02
0.66
33.00
3.00
35219.00
0.02
0.43
21.50
3.00
35248.00
0.02
0.41
20.50
3.00
35299.00
0.03
0.61
20.33
3.00
35332.00
0.06
0.50
8.33
3.00
35571.00
0.01
0.44
44.00
3.00
35590.00
0.02
0.48
24.00
3.00
35612.00
0.02
0.41
20.50
3.00
35646.00
0.03
0.31
10.33
3.00
35691.00
0.02
0.31
15.50
3.00
35963.00
0.02
0.21
10.50
3.00
35983.00
0.03
0.21
7.00
3.00
36012.00
0.01
0.41
41.00
3.00
36320.00
0.03
0.45
15.00
3.00
36362.00
0.01
0.41
41.00
3.00
36376.00
0.03
0.43
14.33
Date
Total Phosphorus (mg/L)
Total Phosphorus (Station CPF0880A)
Date
Total Nitrogen (mg/L)
Total Nitrogen (Station CPF0880A)
Date
N:P Ratio
N:P Ratio (Station CPF0880A)
1.00
30146.00
0.13
4.90
37.69
1.00
30173.00
0.10
0.82
8.20
1.00
30209.00
0.14
1.33
9.50
1.00
30228.00
0.12
1.33
11.08
1.00
30271.00
0.14
1.04
7.43
1.00
30305.00
0.20
1.40
7.00
1.00
30329.00
0.12
1.09
9.08
1.00
30369.00
0.16
1.02
6.38
1.00
30404.00
0.13
0.88
6.77
1.00
30434.00
0.10
0.73
7.30
1.00
30447.00
0.08
0.61
7.63
1.00
30474.00
0.12
1.11
9.25
1.00
30504.00
0.09
0.51
5.67
1.00
30531.00
0.15
0.99
6.60
1.00
30566.00
0.17
0.86
5.06
1.00
30600.00
0.25
2.10
8.40
1.00
30629.00
0.33
2.70
8.18
1.00
30664.00
0.13
1.19
9.15
1.00
30711.00
0.18
1.02
5.67
1.00
30742.00
0.17
0.89
5.24
1.00
30768.00
0.13
0.82
6.31
1.00
30791.00
0.12
0.84
7.00
1.00
30819.00
0.13
0.64
4.92
1.00
30848.00
0.13
0.61
4.69
1.00
30874.00
0.12
0.61
5.08
1.00
30910.00
0.11
0.51
4.64
1.00
30943.00
0.15
0.82
5.47
1.00
30980.00
0.11
0.79
7.18
1.00
31000.00
0.10
1.21
12.10
1.00
31028.00
0.12
1.47
12.25
1.00
31085.00
0.12
1.22
10.17
1.00
31127.00
0.18
1.14
6.33
1.00
31134.00
0.19
1.34
7.05
1.00
31155.00
0.15
1.05
7.00
1.00
31169.00
0.24
0.70
2.92
1.00
31211.00
0.15
0.61
4.07
1.00
31239.00
0.14
0.91
6.50
1.00
31273.00
0.11
0.61
5.55
1.00
31320.00
0.09
0.56
6.22
1.00
31344.00
0.10
0.77
7.70
1.00
31370.00
0.11
0.80
7.27
1.00
31391.00
0.13
0.79
6.08
1.00
31428.00
0.12
1.17
9.75
1.00
31460.00
0.17
1.23
7.24
1.00
31496.00
0.19
0.90
4.74
1.00
31517.00
0.17
0.83
4.88
1.00
31551.00
0.12
0.72
6.00
1.00
31580.00
0.12
0.81
6.75
1.00
31616.00
0.22
1.01
4.59
1.00
31631.00
0.25
1.12
4.48
1.00
31680.00
0.12
0.71
5.92
1.00
31706.00
0.27
1.48
5.48
1.00
31735.00
0.47
4.00
8.51
1.00
31762.00
0.34
3.40
10.00
1.00
31784.00
0.26
2.30
8.85
1.00
31834.00
0.23
1.30
5.65
1.00
31860.00
0.20
0.87
4.35
1.00
31876.00
0.15
0.67
4.47
1.00
31909.00
0.11
0.51
4.64
1.00
31931.00
0.12
0.61
5.08
1.00
32616.00
0.10
0.60
6.00
1.00
32672.00
0.14
0.51
3.64
1.00
32721.00
0.06
0.71
11.83
1.00
32765.00
0.11
0.61
5.55
1.00
32799.00
0.09
0.55
6.11
1.00
32980.00
0.13
0.63
4.85
1.00
33029.00
0.12
0.41
3.42
1.00
33087.00
0.13
0.61
4.69
1.00
33129.00
0.18
0.91
5.06
1.00
33168.00
0.09
0.71
7.89
1.00
33339.00
0.11
0.65
5.91
1.00
33457.00
0.11
0.71
6.46
1.00
33828.00
0.07
0.81
11.57
1.00
34219.00
0.11
0.71
6.46
1.00
34548.00
0.05
0.71
14.20
1.00
34900.00
0.09
0.61
6.78
1.00
35207.00
0.07
0.65
9.29
1.00
35219.00
0.08
0.75
9.38
1.00
35248.00
0.07
0.51
7.29
1.00
35299.00
0.08
0.51
6.38
1.00
35332.00
0.12
0.61
5.08
1.00
35571.00
0.12
0.41
3.42
1.00
35590.00
0.08
0.73
9.13
1.00
35612.00
0.07
0.51
7.29
1.00
35646.00
0.09
0.41
4.56
1.00
35691.00
0.05
0.01
0.20
1.00
35963.00
0.04
0.31
7.75
1.00
35983.00
0.07
0.21
3.00
1.00
36012.00
0.06
0.51
8.50
1.00
36320.00
0.06
0.45
7.50
1.00
36362.00
0.06
0.51
8.50
1.00
36376.00
0.09
0.61
6.78
Date
Total Phosphorus (mg/L)
Total Phosphorus (Station CPF086C)
Date
Total Nitrogen (mg/L)
Total Nitrogen (Station CPF086C)
Date
N:P Ratio
N:P Ratio (Station CPF086C)
4.00
30531.00
0.17
0.90
5.29
4.00
30566.00
0.42
1.21
2.88
4.00
30600.00
0.28
1.66
5.93
4.00
30629.00
0.18
1.50
8.33
4.00
30664.00
0.01
1.07
107.00
4.00
30711.00
0.18
1.01
5.61
4.00
30146.00
0.11
15.80
143.64
4.00
30173.00
0.17
0.99
5.82
4.00
30209.00
0.06
0.63
10.50
4.00
30228.00
0.10
0.73
7.30
4.00
30271.00
0.17
1.06
6.24
4.00
30305.00
0.26
1.09
4.19
4.00
30329.00
0.25
1.35
5.40
4.00
30369.00
0.15
0.92
6.13
4.00
30404.00
0.23
1.20
5.22
4.00
30434.00
0.15
0.72
4.80
4.00
30447.00
0.19
1.10
5.79
4.00
30474.00
0.17
1.00
5.88
4.00
30504.00
0.30
1.50
5.00
4.00
30531.00
0.25
0.86
3.44
4.00
30566.00
0.12
0.61
5.08
4.00
30600.00
0.12
1.07
8.92
4.00
30629.00
0.16
1.38
8.63
4.00
30664.00
0.23
1.09
4.74
4.00
30711.00
0.20
1.14
5.70
4.00
30742.00
0.22
1.05
4.77
4.00
30768.00
0.19
1.06
5.58
4.00
30791.00
0.21
0.93
4.43
4.00
30819.00
0.20
1.20
6.00
4.00
30847.00
0.14
0.53
3.79
4.00
30874.00
0.19
0.90
4.74
4.00
30910.00
0.11
0.41
3.73
4.00
30943.00
0.10
0.92
9.20
4.00
30980.00
0.31
0.85
2.74
4.00
31000.00
0.18
1.03
5.72
4.00
31028.00
0.25
1.26
5.04
4.00
31089.00
0.16
1.24
7.75
4.00
31134.00
0.17
1.18
6.94
4.00
31155.00
0.22
0.93
4.23
4.00
31169.00
0.22
0.94
4.27
4.00
31211.00
0.30
0.99
3.30
4.00
31219.00
0.29
0.81
2.79
4.00
31222.00
0.23
0.81
3.52
4.00
31239.00
0.30
0.93
3.10
4.00
31273.00
0.11
0.61
5.55
4.00
31320.00
0.09
0.78
8.67
4.00
31344.00
0.13
0.78
6.00
4.00
31370.00
0.26
1.22
4.69
4.00
31391.00
0.09
0.98
10.89
4.00
31428.00
0.42
2.00
4.76
4.00
31460.00
0.27
1.11
4.11
4.00
31496.00
0.17
1.03
6.06
4.00
31517.00
0.14
0.71
5.07
4.00
31551.00
0.23
0.42
1.83
4.00
31580.00
0.33
0.93
2.82
4.00
31616.00
0.35
0.81
2.31
4.00
31631.00
0.28
0.61
2.18
4.00
31680.00
0.11
0.51
4.64
4.00
31706.00
0.14
0.94
6.71
4.00
31735.00
0.20
1.38
6.90
4.00
31762.00
0.43
1.60
3.72
4.00
31784.00
0.29
1.50
5.17
4.00
31834.00
0.16
0.97
6.06
4.00
31860.00
0.23
1.13
4.91
4.00
31876.00
0.19
0.90
4.74
4.00
31909.00
0.20
0.84
4.20
4.00
31931.00
0.19
0.61
3.21
4.00
32014.00
0.14
0.85
6.07
4.00
32044.00
0.22
0.96
4.36
4.00
32065.00
0.16
1.12
7.00
4.00
32254.00
0.23
1.19
5.17
4.00
32296.00
0.18
0.93
5.17
4.00
32364.00
0.24
0.71
2.96
4.00
32413.00
0.20
1.23
6.15
4.00
32440.00
0.12
1.04
8.67
4.00
32616.00
0.14
1.19
8.50
4.00
32672.00
0.16
1.03
6.44
4.00
32721.00
0.14
0.82
5.86
4.00
32765.00
0.11
0.61
5.55
4.00
32799.00
0.07
0.72
10.29
4.00
32980.00
0.16
0.99
6.19
4.00
33029.00
0.08
0.91
11.38
4.00
33087.00
0.08
0.86
10.75
4.00
33129.00
0.09
0.91
10.11
4.00
33168.00
0.14
0.98
7.00
4.00
33339.00
0.08
1.02
12.75
4.00
33457.00
0.06
0.71
11.83
4.00
33828.00
0.06
0.65
10.83
4.00
34219.00
0.07
0.70
10.00
4.00
34548.00
0.16
0.79
4.94
4.00
34900.00
0.11
0.63
5.73
4.00
35207.00
0.07
0.65
9.29
4.00
35219.00
0.07
0.94
13.43
4.00
35248.00
0.05
0.51
10.20
4.00
35299.00
0.05
0.61
12.20
4.00
35332.00
0.07
0.60
8.57
4.00
35571.00
0.10
1.04
10.40
4.00
35590.00
0.06
0.85
14.17
4.00
35612.00
0.06
0.88
14.67
4.00
35646.00
0.11
0.61
5.55
4.00
35691.00
0.06
0.21
3.50
4.00
35963.00
0.07
0.85
12.14
4.00
35983.00
0.07
0.42
6.00
4.00
36012.00
0.05
0.52
10.40
4.00
36320.00
0.07
0.55
7.86
4.00
36362.00
0.17
0.61
3.59
4.00
30146.00
0.11
10.70
97.27
4.00
30173.00
0.10
0.71
7.10
4.00
30209.00
0.05
0.61
12.20
4.00
30228.00
0.10
0.83
8.30
4.00
30271.00
0.11
0.95
8.64
4.00
30305.00
0.11
1.04
9.46
4.00
30329.00
0.10
1.14
11.40
4.00
30369.00
0.14
0.97
6.93
4.00
30404.00
0.08
1.03
12.88
4.00
30434.00
0.08
1.13
14.13
4.00
30447.00
0.13
0.94
7.23
4.00
30474.00
0.09
0.72
8.00
4.00
30504.00
0.12
0.62
5.17
4.00
30531.00
0.12
0.71
5.92
4.00
30566.00
0.06
0.43
7.17
4.00
30600.00
0.07
0.83
11.86
4.00
30629.00
0.13
1.32
10.15
4.00
30664.00
0.18
1.03
5.72
4.00
30711.00
0.09
1.12
12.44
4.00
30742.00
0.16
1.16
7.25
4.00
30768.00
0.18
0.92
5.11
4.00
30791.00
0.15
0.78
5.20
4.00
30819.00
0.17
1.25
7.35
4.00
30847.00
0.09
0.61
6.78
4.00
30874.00
0.14
0.82
5.86
4.00
30910.00
0.06
0.41
6.83
4.00
30943.00
0.08
0.78
9.75
4.00
30980.00
0.05
0.57
11.40
4.00
31000.00
0.11
1.03
9.36
4.00
31028.00
0.23
1.18
5.13
4.00
31089.00
0.13
1.02
7.85
4.00
31134.00
0.14
1.24
8.86
4.00
31155.00
0.11
0.81
7.36
4.00
31169.00
0.15
0.85
5.67
4.00
31211.00
0.21
1.03
4.91
4.00
31219.00
0.21
0.71
3.38
4.00
31222.00
0.16
0.61
3.81
4.00
31239.00
0.10
0.61
6.10
4.00
31273.00
0.09
0.61
6.78
4.00
31320.00
0.05
0.64
12.80
4.00
31344.00
0.07
0.92
13.14
4.00
31370.00
0.10
0.97
9.70
4.00
31391.00
0.08
1.11
13.88
4.00
31428.00
0.19
1.50
7.90
4.00
31460.00
0.27
1.30
4.82
4.00
31496.00
0.13
0.96
7.39
4.00
31517.00
0.10
0.59
5.90
4.00
31551.00
0.18
0.91
5.06
4.00
31580.00
0.24
0.90
3.75
4.00
31616.00
0.10
0.51
5.10
4.00
31631.00
0.10
0.61
6.10
4.00
31680.00
0.10
0.51
5.10
4.00
31706.00
0.11
1.11
10.09
4.00
31735.00
0.19
1.40
7.37
4.00
31762.00
0.35
1.60
4.57
4.00
31784.00
0.22
1.55
7.05
4.00
31834.00
0.19
1.18
6.21
4.00
31860.00
0.15
0.96
6.40
4.00
31876.00
0.17
0.93
5.47
4.00
31909.00
0.10
0.73
7.30
4.00
31931.00
0.20
0.61
3.05
4.00
32014.00
0.06
0.81
13.50
4.00
32044.00
0.14
0.78
5.57
4.00
32065.00
0.10
0.98
9.80
4.00
32254.00
0.16
1.05
6.56
4.00
32296.00
0.11
0.63
5.73
4.00
32364.00
0.07
0.61
8.71
4.00
32413.00
0.10
0.88
8.80
4.00
32440.00
0.11
1.10
10.00
4.00
32616.00
0.08
0.80
10.00
4.00
32672.00
0.14
0.91
6.50
4.00
32721.00
0.10
0.80
8.00
4.00
32765.00
0.06
0.51
8.50
4.00
32799.00
0.06
0.68
11.33
4.00
32980.00
0.10
0.91
9.10
4.00
33029.00
0.06
0.79
13.17
4.00
33087.00
0.06
0.51
8.50
4.00
33129.00
0.05
0.61
12.20
4.00
33168.00
0.13
1.08
8.31
4.00
33339.00
0.06
0.85
14.17
4.00
33457.00
0.03
0.51
17.00
4.00
33828.00
0.05
0.64
12.80
4.00
34219.00
0.04
0.59
14.75
4.00
34548.00
0.04
0.61
15.25
4.00
34900.00
0.09
0.61
6.78
4.00
35207.00
0.04
0.58
14.50
4.00
35219.00
0.04
0.67
16.75
4.00
35248.00
0.04
0.31
7.75
4.00
35299.00
0.04
0.71
17.75
4.00
35332.00
0.06
0.55
9.17
4.00
35571.00
0.07
0.45
6.43
4.00
35590.00
0.04
0.65
16.25
4.00
35612.00
0.04
0.52
13.00
4.00
35646.00
0.09
0.41
4.56
4.00
35691.00
0.06
0.41
6.83
4.00
35963.00
0.05
0.77
15.40
4.00
35983.00
0.05
0.41
8.20
4.00
36012.00
0.02
0.31
15.50
4.00
36320.00
0.03
0.45
15.00
4.00
36362.00
0.06
0.51
8.50
4.00
36376.00
0.07
0.71
10.14
4.00
30847.00
0.13
0.95
7.31
1.00
30146.00
0.08
9.10
113.75
1.00
30173.00
0.09
0.61
6.78
1.00
30209.00
0.09
1.51
16.78
1.00
30228.00
0.12
1.23
10.25
1.00
30271.00
0.11
0.96
8.73
1.00
30305.00
0.18
1.00
5.56
1.00
30329.00
0.18
1.19
6.61
1.00
30369.00
0.13
0.65
5.00
1.00
30404.00
0.14
0.71
5.07
1.00
30434.00
0.12
0.78
6.50
1.00
30447.00
0.07
0.61
8.71
1.00
30474.00
0.10
0.61
6.10
1.00
30504.00
0.15
0.71
4.73
1.00
30531.00
0.12
0.71
5.92
1.00
30566.00
0.24
1.82
7.58
1.00
30600.00
0.22
1.90
8.64
1.00
30629.00
0.28
2.10
7.50
1.00
30664.00
0.21
1.04
4.95
1.00
30711.00
0.16
0.79
4.94
1.00
30742.00
0.16
0.88
5.50
1.00
30768.00
0.15
0.94
6.27
1.00
30791.00
0.13
0.80
6.15
1.00
30819.00
0.13
0.41
3.15
1.00
30848.00
0.17
0.61
3.59
1.00
30874.00
0.13
0.71
5.46
1.00
30910.00
0.09
0.51
5.67
1.00
30943.00
0.13
0.61
4.69
1.00
30980.00
0.09
0.96
10.67
1.00
31000.00
0.10
1.36
13.60
1.00
31028.00
0.20
1.82
9.10
1.00
31085.00
0.16
1.24
7.75
1.00
31127.00
0.18
1.19
6.61
1.00
31134.00
0.14
1.13
8.07
1.00
31155.00
0.14
0.86
6.14
1.00
31169.00
0.13
1.01
7.77
1.00
31211.00
0.16
0.92
5.75
1.00
31239.00
0.16
0.91
5.69
1.00
31273.00
0.10
0.61
6.10
1.00
31320.00
0.11
0.75
6.82
1.00
31344.00
0.10
0.76
7.60
1.00
31370.00
0.11
0.99
9.00
1.00
31391.00
0.16
0.96
6.00
1.00
31428.00
0.12
1.01
8.42
1.00
31460.00
0.26
1.90
7.31
1.00
31496.00
0.16
1.05
6.56
1.00
31517.00
0.14
0.63
4.50
1.00
31551.00
0.14
0.92
6.57
1.00
31580.00
0.14
0.81
5.79
1.00
31616.00
0.27
1.18
4.37
1.00
31631.00
0.29
1.03
3.55
1.00
31680.00
0.17
1.01
5.94
1.00
31706.00
0.30
2.12
7.07
1.00
31735.00
0.88
5.20
5.91
1.00
31762.00
0.55
4.10
7.46
1.00
31784.00
0.24
2.30
9.58
1.00
31834.00
0.24
1.10
4.58
1.00
31860.00
0.17
0.67
3.94
1.00
31876.00
0.15
0.62
4.13
1.00
31909.00
0.12
0.51
4.25
1.00
31931.00
0.12
0.61
5.08
1.00
32616.00
0.11
0.59
5.36
1.00
32672.00
0.14
0.71
5.07
1.00
32721.00
0.11
0.61
5.55
1.00
32765.00
0.09
0.81
9.00
1.00
32799.00
0.10
0.88
8.80
1.00
32980.00
0.12
0.66
5.50
1.00
33029.00
0.12
0.61
5.08
1.00
33087.00
0.13
0.81
6.23
1.00
33129.00
0.17
1.04
6.12
1.00
33168.00
0.10
0.81
8.10
1.00
33339.00
0.08
0.77
9.63
1.00
33457.00
0.10
0.71
7.10
1.00
33828.00
0.07
0.71
10.14
1.00
34219.00
0.14
0.61
4.36
1.00
34548.00
0.05
0.61
12.20
1.00
34900.00
0.11
0.61
5.55
1.00
35207.00
0.06
0.73
12.17
1.00
35219.00
0.11
0.45
4.09
1.00
35248.00
0.10
0.61
6.10
1.00
35299.00
0.09
0.31
3.44
1.00
35332.00
0.22
0.61
2.77
1.00
35571.00
0.01
0.31
31.00
1.00
35590.00
0.09
0.75
8.33
1.00
35612.00
0.09
0.61
6.78
1.00
35646.00
0.09
0.71
7.89
1.00
35691.00
0.07
0.41
5.86
1.00
35963.00
0.09
0.31
3.44
1.00
35983.00
0.10
0.31
3.10
1.00
36012.00
0.11
0.31
2.82
1.00
36320.00
0.07
0.53
7.57
1.00
36362.00
0.08
0.61
7.63
1.00
36376.00
0.09
1.41
15.67
1.00
30146.00
0.13
4.90
37.69
1.00
30173.00
0.10
0.82
8.20
1.00
30209.00
0.14
1.33
9.50
1.00
30228.00
0.12
1.33
11.08
1.00
30271.00
0.14
1.04
7.43
1.00
30305.00
0.20
1.40
7.00
1.00
30329.00
0.12
1.09
9.08
1.00
30369.00
0.16
1.02
6.38
1.00
30404.00
0.13
0.88
6.77
1.00
30434.00
0.10
0.73
7.30
1.00
30447.00
0.08
0.61
7.63
1.00
30474.00
0.12
1.11
9.25
1.00
30504.00
0.09
0.51
5.67
1.00
30531.00
0.15
0.99
6.60
1.00
30566.00
0.17
0.86
5.06
1.00
30600.00
0.25
2.10
8.40
1.00
30629.00
0.33
2.70
8.18
1.00
30664.00
0.13
1.19
9.15
1.00
30711.00
0.18
1.02
5.67
1.00
30742.00
0.17
0.89
5.24
1.00
30768.00
0.13
0.82
6.31
1.00
30791.00
0.12
0.84
7.00
1.00
30819.00
0.13
0.64
4.92
1.00
30848.00
0.13
0.61
4.69
1.00
30874.00
0.12
0.61
5.08
1.00
30910.00
0.11
0.51
4.64
1.00
30943.00
0.15
0.82
5.47
1.00
30980.00
0.11
0.79
7.18
1.00
31000.00
0.10
1.21
12.10
1.00
31028.00
0.12
1.47
12.25
1.00
31085.00
0.12
1.22
10.17
1.00
31127.00
0.18
1.14
6.33
1.00
31134.00
0.19
1.34
7.05
1.00
31155.00
0.15
1.05
7.00
1.00
31169.00
0.24
0.70
2.92
1.00
31211.00
0.15
0.61
4.07
1.00
31239.00
0.14
0.91
6.50
1.00
31273.00
0.11
0.61
5.55
1.00
31320.00
0.09
0.56
6.22
1.00
31344.00
0.10
0.77
7.70
1.00
31370.00
0.11
0.80
7.27
1.00
31391.00
0.13
0.79
6.08
1.00
31428.00
0.12
1.17
9.75
1.00
31460.00
0.17
1.23
7.24
1.00
31496.00
0.19
0.90
4.74
1.00
31517.00
0.17
0.83
4.88
1.00
31551.00
0.12
0.72
6.00
1.00
31580.00
0.12
0.81
6.75
1.00
31616.00
0.22
1.01
4.59
1.00
31631.00
0.25
1.12
4.48
1.00
31680.00
0.12
0.71
5.92
1.00
31706.00
0.27
1.48
5.48
1.00
31735.00
0.47
4.00
8.51
1.00
31762.00
0.34
3.40
10.00
1.00
31784.00
0.26
2.30
8.85
1.00
31834.00
0.23
1.30
5.65
1.00
31860.00
0.20
0.87
4.35
1.00
31876.00
0.15
0.67
4.47
1.00
31909.00
0.11
0.51
4.64
1.00
31931.00
0.12
0.61
5.08
1.00
32616.00
0.10
0.60
6.00
1.00
32672.00
0.14
0.51
3.64
1.00
32721.00
0.06
0.71
11.83
1.00
32765.00
0.11
0.61
5.55
1.00
32799.00
0.09
0.55
6.11
1.00
32980.00
0.13
0.63
4.85
1.00
33029.00
0.12
0.41
3.42
1.00
33087.00
0.13
0.61
4.69
1.00
33129.00
0.18
0.91
5.06
1.00
33168.00
0.09
0.71
7.89
1.00
33339.00
0.11
0.65
5.91
1.00
33457.00
0.11
0.71
6.46
1.00
33828.00
0.07
0.81
11.57
1.00
34219.00
0.11
0.71
6.46
1.00
34548.00
0.05
0.71
14.20
1.00
34900.00
0.09
0.61
6.78
1.00
35207.00
0.07
0.65
9.29
1.00
35219.00
0.08
0.75
9.38
1.00
35248.00
0.07
0.51
7.29
1.00
35299.00
0.08
0.51
6.38
1.00
35332.00
0.12
0.61
5.08
1.00
35571.00
0.12
0.41
3.42
1.00
35590.00
0.08
0.73
9.13
1.00
35612.00
0.07
0.51
7.29
1.00
35646.00
0.09
0.41
4.56
1.00
35691.00
0.05
0.01
0.20
1.00
35963.00
0.04
0.31
7.75
1.00
35983.00
0.07
0.21
3.00
1.00
36012.00
0.06
0.51
8.50
1.00
36320.00
0.06
0.45
7.50
1.00
36362.00
0.06
0.51
8.50
1.00
36376.00
0.09
0.61
6.78
1.00
30173.00
0.08
0.61
7.63
1.00
30209.00
0.07
1.01
14.43
1.00
30228.00
0.08
0.93
11.63
1.00
30271.00
0.09
0.83
9.22
1.00
30305.00
0.09
1.15
12.78
1.00
30329.00
0.09
1.13
12.56
1.00
30369.00
0.13
0.94
7.23
1.00
30404.00
0.11
0.77
7.00
1.00
30434.00
0.08
0.75
9.38
1.00
30447.00
0.07
0.61
8.71
1.00
30474.00
0.07
0.71
10.14
1.00
30504.00
0.06
0.51
8.50
1.00
30531.00
0.08
0.71
8.88
1.00
30566.00
0.08
0.96
12.00
1.00
30600.00
0.10
1.29
12.90
1.00
30629.00
0.10
1.82
18.20
1.00
30664.00
0.13
1.32
10.15
1.00
30711.00
0.15
0.82
5.47
1.00
30742.00
0.16
1.02
6.38
1.00
30768.00
0.11
0.77
7.00
1.00
30791.00
0.11
0.81
7.36
1.00
30819.00
0.07
0.41
5.86
1.00
30848.00
0.08
0.51
6.38
1.00
30874.00
0.09
0.81
9.00
1.00
30910.00
0.07
0.51
7.29
1.00
30943.00
0.13
0.71
5.46
1.00
30980.00
0.06
0.77
12.83
1.00
31000.00
0.09
1.10
12.22
1.00
31028.00
0.07
1.25
17.86
1.00
31085.00
0.09
1.08
12.00
1.00
31127.00
0.14
1.10
7.86
1.00
31134.00
0.10
1.06
10.60
1.00
31155.00
0.11
0.72
6.55
1.00
31169.00

0.10
0.61
6.10
1.00
31211.00
0.10
0.71
7.10
1.00
31239.00
0.15
0.71
4.73
1.00
31273.00
0.12
0.81
6.75
1.00
31320.00
0.07
0.55
7.86
1.00
31344.00
0.08
0.99
12.38
1.00
31370.00
0.08
0.76
9.50
1.00
31391.00
0.13
0.81
6.23
1.00
31428.00
0.17
1.18
6.94
1.00
31460.00
0.16
1.32
8.25
1.00
31496.00
0.18
0.64
3.56
1.00
31517.00
0.13
0.81
6.23
1.00
31551.00
0.10
0.72
7.20
1.00
31580.00
0.06
0.71
11.83
1.00
31616.00
0.15
0.82
5.47
1.00
31631.00
0.12
0.82
6.83
1.00
31680.00
0.10
0.61
6.10
1.00
31706.00
0.16
0.80
5.00
1.00
31735.00
0.24
2.50
10.42
1.00
31762.00
0.20
2.70
13.50
1.00
31784.00
0.25
2.30
9.20
1.00
31834.00
0.20
1.22
6.10
1.00
31860.00
0.20
0.87
4.35
1.00
31876.00
0.13
0.66
5.08
1.00
31909.00
0.08
0.41
5.13
1.00
31931.00
0.09
0.51
5.67
1.00
32014.00
0.12
0.71
5.92
1.00
32044.00
0.12
1.11
9.25
1.00
32065.00
0.14
0.61
4.36
1.00
32254.00
0.13
0.74
5.69
1.00
32296.00
0.08
0.61
7.63
1.00
32364.00
0.12
0.71
5.92
1.00
32413.00
0.19
1.27
6.68
1.00
32440.00
0.13
1.04
8.00
1.00
32616.00
0.08
0.65
8.13
1.00
32672.00
0.08
0.51
6.38
1.00
32721.00
0.07
0.61
8.71
1.00
32765.00
0.09
0.61
6.78
1.00
32799.00
0.08
0.56
7.00
1.00
32980.00
0.11
0.67
6.09
1.00
33029.00
0.11
0.61
5.55
1.00
33087.00
0.14
0.71
5.07
1.00
33129.00
0.09
0.81
9.00
1.00
33168.00
0.08
0.81
10.13
1.00
33339.00
0.07
0.69
9.86
1.00
33457.00
0.08
0.71
8.88
1.00
33828.00
0.09
0.81
9.00
1.00
34219.00
0.06
0.71
11.83
1.00
34548.00
0.04
0.41
10.25
1.00
34900.00
0.09
0.71
7.89
1.00
35207.00
0.06
0.76
12.67
1.00
35219.00
0.08
0.40
5.00
1.00
35248.00
0.06
0.51
8.50
1.00
35299.00
0.06
0.41
6.83
1.00
35332.00
0.11
0.51
4.64
1.00
35571.00
0.05
0.41
8.20
1.00
35590.00
0.12
0.51
4.25
1.00
35612.00
0.05
0.41
8.20
1.00
35646.00
0.07
0.41
5.86
1.00
35691.00
0.07
0.61
8.71
1.00
35963.00
0.06
0.31
5.17
1.00
35983.00
0.06
0.31
5.17
1.00
36012.00
0.03
0.51
17.00
1.00
36320.00
0.04
0.43
10.75
1.00
36362.00
0.05
0.64
12.80
1.00
36376.00
0.09
0.71
7.89
2.00
30146.00
0.07
4.80
68.57
2.00
30173.00
0.06
0.61
10.17
2.00
30209.00
0.06
0.71
11.83
2.00
30228.00
0.09
1.03
11.44
2.00
30271.00
0.06
0.67
11.17
2.00
30305.00
0.09
1.12
12.44
2.00
30329.00
0.08
1.01
12.63
2.00
30369.00
0.11
1.28
11.64
2.00
30404.00
0.10
0.80
8.00
2.00
30434.00
0.07
0.67
9.57
2.00
30447.00
0.05
0.62
12.40
2.00
30474.00
0.05
0.41
8.20
2.00
30504.00
0.05
0.31
6.20
2.00
30531.00
0.05
0.51
10.20
2.00
30566.00
0.05
0.61
12.20
2.00
30600.00
0.08
0.84
10.50
2.00
30629.00
0.08
1.42
17.75
2.00
30664.00
0.07
1.24
17.71
2.00
30711.00
0.24
0.90
3.75
2.00
30742.00
0.11
0.79
7.18
2.00
30768.00
0.12
0.71
5.92
2.00
30791.00
0.08
0.66
8.25
2.00
30819.00
0.06
0.47
7.83
2.00
30848.00
0.05
0.41
8.20
2.00
30874.00
0.05
0.61
12.20
2.00
30910.00
0.06
0.51
8.50
2.00
30943.00
0.08
0.51
6.38
2.00
30980.00
0.06
0.64
10.67
2.00
31000.00
0.06
0.84
14.00
2.00
31028.00
0.04
0.68
17.00
2.00
31085.00
0.05
0.88
17.60
2.00
30209.00
0.05
0.93
18.60
2.00
30228.00
0.11
1.03
9.36
2.00
30271.00
0.05
0.63
12.60
2.00
30305.00
0.05
0.88
17.60
2.00
30329.00
0.05
0.96
19.20
2.00
30369.00
0.10
0.89
8.90
2.00
30404.00
0.09
0.94
10.44
2.00
30434.00
0.06
0.71
11.83
2.00
30447.00
0.09
0.97
10.78
2.00
30474.00
0.04
0.41
10.25
2.00
30504.00
0.04
0.31
7.75
2.00
30531.00
0.04
0.51
12.75
2.00
30566.00
0.05
0.61
12.20
2.00
30600.00
0.06
0.89
14.83
2.00
30629.00
0.08
1.03
12.88
2.00
30664.00
0.05
1.03
20.60
2.00
30711.00
0.22
0.92
4.18
2.00
30742.00
0.09
0.92
10.22
2.00
30768.00
0.09
0.56
6.22
2.00
30791.00
0.07
0.77
11.00
2.00
30819.00
0.05
0.51
10.20
2.00
30848.00
0.05
0.41
8.20
2.00
30874.00
0.05
0.41
8.20
2.00
30910.00
0.05
0.51
10.20
2.00
30943.00
0.05
0.81
16.20
2.00
30980.00
0.04
0.46
11.50
2.00
31000.00
0.05
0.72
14.40
2.00
31028.00
0.04
0.66
16.50
2.00
31085.00
0.07
0.69
9.86
2.00
31134.00
0.05
0.77
15.40
2.00
31155.00
0.05
0.53
10.60
2.00
31169.00
0.05
0.44
8.80
2.00
31211.00
0.04
0.51
12.75
2.00
31239.00
0.04
0.61
15.25
2.00
31273.00
0.04
0.61
15.25
2.00
31320.00
0.04
0.56
14.00
2.00
31344.00
0.05
0.62
12.40
2.00
31370.00
0.05
0.82
16.40
2.00
31391.00
0.06
0.86
14.33
2.00
31428.00
0.05
0.85
17.00
2.00
31460.00
0.06
0.90
15.00
2.00
31496.00
0.08
1.08
13.50
2.00
31517.00
0.05
0.64
12.80
2.00
31551.00
0.05
0.62
12.40
2.00
31580.00
0.03
0.51
17.00
2.00
31616.00
0.05
0.41
8.20
2.00
31631.00
0.05
0.41
8.20
2.00
31680.00
0.05
0.09
1.80
2.00
31706.00
0.06
0.77
12.83
2.00
31735.00
0.03
1.12
37.33
2.00
31762.00
0.04
1.17
29.25
2.00
31784.00
0.05
1.18
23.60
2.00
31834.00
0.11
1.35
12.27
2.00
31860.00
0.11
0.65
5.91
2.00
31876.00
0.07
0.55
7.86
2.00
31909.00
0.04
0.43
10.75
2.00
31931.00
0.04
0.41
10.25
2.00
32014.00
0.06
0.61
10.17
2.00
32044.00
0.04
0.61
15.25
2.00
32065.00
0.07
0.73
10.43
2.00
32254.00
0.06
0.79
13.17
2.00
32296.00
0.04
0.51
12.75
2.00
32364.00
0.04
0.51
12.75
2.00
32413.00
0.07
0.81
11.57
2.00
32440.00
0.07
0.61
8.71
2.00
32616.00
0.07
0.62
8.86
2.00
32672.00
0.05
0.41
8.20
2.00
32721.00
0.05
0.41
8.20
2.00
32765.00
0.05
0.51
10.20
2.00
32799.00
0.06
0.52
8.67
2.00
32980.00
0.07
0.56
8.00
2.00
33029.00
0.05
0.54
10.80
2.00
33087.00
0.04
0.51
12.75
2.00
33129.00
0.04
0.61
15.25
2.00
33168.00
0.05
0.64
12.80
2.00
33339.00
0.04
0.65
16.25
2.00
33457.00
0.02
0.51
25.50
2.00
33828.00
0.03
0.61
20.33
2.00
34219.00
0.04
0.51
12.75
2.00
34548.00
0.01
0.41
41.00
2.00
34900.00
0.04
0.51
12.75
2.00
35207.00
0.03
0.79
26.33
2.00
35219.00
0.04
0.52
13.00
2.00
35248.00
0.02
0.41
20.50
2.00
35299.00
0.03
0.51
17.00
2.00
35332.00
0.08
0.51
6.38
2.00
35571.00
0.04
0.31
7.75
2.00
35590.00
0.07
0.46
6.57
2.00
35612.00
0.02
0.61
30.50
2.00
35646.00
0.04
0.41
10.25
2.00
35691.00
0.02
0.21
10.50
2.00
35963.00
0.03
0.21
7.00
2.00
35983.00
0.03
0.31
10.33
2.00
36012.00
0.01
0.31
31.00
2.00
36320.00
0.02
0.43
21.50
2.00
36362.00
0.01
0.63
63.00
2.00
36376.00
0.04
0.71
17.75
2.00
30173.00
0.04
0.51
12.75
2.00
30209.00
0.04
0.81
20.25
2.00
30228.00
0.05
0.63
12.60
2.00
30271.00
0.05
0.60
12.00
2.00
30305.00
0.06
0.89
14.83
2.00
30329.00
0.05
0.98
19.60
2.00
30369.00
0.08
1.00
12.50
2.00
30404.00
0.08
0.90
11.25
2.00
30434.00
0.07
0.73
10.43
2.00
30447.00
0.04
0.72
18.00
2.00
30474.00
0.04
0.41
10.25
2.00
30504.00
0.03
0.31
10.33
2.00
30531.00
0.03
0.51
17.00
2.00
30566.00
0.04
0.41
10.25
2.00
30600.00
0.06
0.65
10.83
2.00
30629.00
0.05
0.93
18.60
2.00
30664.00
0.05
1.09
21.80
2.00
30711.00
0.20
0.92
4.60
2.00
30742.00
0.08
0.91
11.38
2.00
30768.00
0.07
0.59
8.43
2.00
30791.00
0.06
0.67
11.17
2.00
30819.00
0.05
0.51
10.20
2.00
30848.00
0.05
0.41
8.20
2.00
30874.00
0.05
0.41
8.20
2.00
30910.00
0.05
0.31
6.20
2.00
30943.00
0.04
0.51
12.75
2.00
30980.00
0.04
0.47
11.75
2.00
31000.00
0.04
0.72
18.00
2.00
31028.00
0.05
0.67
13.40
2.00
31089.00
0.05
0.76
15.20
2.00
35207.00
0.02
0.67
33.50
2.00
35219.00
0.02
0.62
31.00
2.00
35248.00
0.04
0.51
12.75
2.00
35299.00
0.02
0.51
25.50
2.00
35332.00
0.04
0.61
15.25
2.00
35571.00
0.05
0.41
8.20
2.00
35590.00
0.06
0.35
5.83
2.00
35612.00
0.02
0.41
20.50
2.00
35646.00
0.03
0.42
14.00
2.00
35691.00
0.02
0.42
21.00
2.00
36320.00
0.02
0.44
22.00
2.00
36362.00
0.01
0.41
41.00
2.00
36376.00
0.04
0.51
12.75
3.00
30146.00
0.06
8.20
136.67
3.00
30173.00
0.05
0.51
10.20
3.00
30209.00
0.04
0.81
20.25
3.00
30228.00
0.06
0.93
15.50
3.00
30271.00
0.05
0.67
13.40
3.00
30305.00
0.05
0.91
18.20
3.00
30329.00
0.04
1.06
26.50
3.00
30369.00
0.05
0.93
18.60
3.00
30404.00
0.06
1.01
16.83
3.00
30434.00
0.06
0.71
11.83
3.00
30447.00
0.08
0.89
11.13
3.00
30474.00
0.04
0.51
12.75
3.00
30504.00
0.04
0.41
10.25
3.00
30531.00
0.04
0.41
10.25
3.00
30566.00
0.04
0.41
10.25
3.00
30600.00
0.06
0.86
14.33
3.00
30629.00
0.05
0.79
15.80
3.00
30664.00
0.07
1.21
17.29
3.00
30711.00
0.20
0.93
4.65
3.00
30742.00
0.09
0.91
10.11
3.00
30768.00
0.08
0.69
8.63
3.00
30791.00
0.07
0.69
9.86
3.00
30819.00
0.05
0.55
11.00
3.00
30847.00
3.00
30848.00
0.05
0.31
6.20
3.00
30874.00
0.06
0.41
6.83
3.00
30910.00
0.05
0.41
8.20
3.00
30943.00
0.04
0.51
12.75
3.00
30980.00
0.04
0.58
14.50
3.00
31000.00
0.04
0.86
21.50
3.00
31028.00
0.04
0.91
22.75
3.00
31089.00
0.07
0.79
11.29
3.00
31134.00
0.05
0.76
15.20
3.00
31155.00
0.05
0.67
13.40
3.00
31169.00
0.05
0.31
6.20
3.00
31211.00
0.04
0.51
12.75
3.00
31239.00
0.03
0.51
17.00
3.00
31273.00
0.04
0.41
10.25
3.00
31320.00
0.04
0.59
14.75
3.00
31344.00
0.04
0.62
15.50
3.00
31370.00
0.06
0.87
14.50
3.00
31391.00
0.09
0.96
10.67
3.00
31428.00
0.04
0.98
24.50
3.00
31460.00
0.04
0.99
24.75
3.00
31496.00
0.05
0.84
16.80
3.00
31517.00
0.04
0.54
13.50
3.00
31551.00
0.06
0.81
13.50
3.00
31580.00
0.03
0.51
17.00
3.00
31616.00
0.05
0.41
8.20
3.00
31631.00
0.04
0.61
15.25
3.00
31680.00
0.04
0.51
12.75
3.00
31706.00
0.06
0.74
12.33
3.00
31735.00
0.03
1.01
33.67
3.00
31762.00
0.04
1.25
31.25
3.00
31784.00
0.10
1.25
12.50
3.00
31834.00
0.09
1.14
12.67
3.00
31860.00
0.10
0.79
7.90
3.00
31876.00
0.06
0.63
10.50
3.00
31909.00
0.05
0.48
9.60
3.00
31931.00
0.04
0.41
10.25
3.00
32413.00
3.00
35207.00
0.02
0.67
33.50
3.00
35219.00
0.03
1.14
38.00
3.00
35248.00
0.03
0.31
10.33
3.00
35299.00
0.02
0.41
20.50
3.00
35332.00
0.04
0.41
10.25
3.00
35571.00
0.04
0.41
10.25
3.00
35590.00
0.06
0.47
7.83
3.00
35612.00
0.02
0.21
10.50
3.00
35646.00
0.03
0.41
13.67
3.00
35691.00
0.01
0.31
31.00
3.00
36320.00
0.02
0.44
22.00
3.00
36362.00
0.01
0.41
41.00
3.00
36376.00
0.03
0.81
27.00
3.00
30146.00
0.05
7.40
148.00
3.00
30173.00
0.04
0.51
12.75
3.00
30209.00
0.03
0.61
20.33
3.00
30228.00
0.09
0.83
9.22
3.00
30271.00
0.06
0.74
12.33
3.00
30305.00
0.08
1.30
16.25
3.00
30329.00
0.05
0.88
17.60
3.00
30369.00
0.06
0.82
13.67
3.00
30404.00
0.06
1.01
16.83
3.00
30434.00
0.07
0.80
11.43
3.00
30447.00
0.05
0.72
14.40
3.00
30474.00
0.04
0.44
11.00
3.00
30504.00
0.04
0.41
10.25
3.00
30531.00
0.06
0.41
6.83
3.00
30566.00
0.04
0.41
10.25
3.00
30600.00
0.06
0.71
11.83
3.00
30629.00
0.06
0.86
14.33
3.00
30664.00
0.11
1.14
10.36
3.00
30711.00
0.07
0.98
14.00
3.00
30742.00
0.11
0.94
8.55
3.00
30768.00
0.08
0.81
10.13
3.00
30791.00
0.08
0.71
8.88
3.00
30819.00
0.05
0.65
13.00
3.00
30848.00
0.06
0.41
6.83
3.00
30874.00
0.12
0.41
3.42
3.00
30910.00
0.05
0.51
10.20
3.00
30943.00
0.05
0.61
12.20
3.00
30980.00
0.03
0.48
16.00
3.00
31000.00
0.04
0.61
15.25
3.00
31028.00
0.07
0.81
11.57
3.00
31089.00
0.08
0.82
10.25
3.00
31134.00
0.06
0.81
13.50
3.00
31155.00
0.06
0.68
11.33
3.00
31169.00
0.05
0.43
8.60
3.00
31211.00
0.04
0.51
12.75
3.00
31222.00
0.04
0.65
16.25
3.00
31239.00
0.04
0.51
12.75
3.00
31273.00
0.04
0.61
15.25
3.00
31320.00
0.04
0.60
15.00
3.00
31344.00
0.05
0.65
13.00
3.00
31370.00
0.08
0.89
11.13
3.00
31391.00
0.06
0.99
16.50
3.00
31428.00
0.05
1.01
20.20
3.00
31460.00
0.05
0.93
18.60
3.00
31496.00
0.06
0.82
13.67
3.00
31517.00
0.05
0.49
9.80
3.00
31551.00
0.05
0.62
12.40
3.00
31580.00
0.03
0.51
17.00
3.00
31616.00
0.04
0.41
10.25
3.00
31631.00
0.04
0.61
15.25
3.00
31680.00
0.03
0.71
23.67
3.00
31706.00
0.06
0.94
15.67
3.00
31735.00
0.03
1.08
36.00
3.00
31762.00
0.12
1.43
11.92
3.00
31784.00
0.19
1.51
7.95
3.00
31834.00
0.09
1.15
12.78
3.00
31860.00
0.11
0.74
6.73
3.00
31876.00
0.06
0.74
12.33
3.00
31909.00
0.04
0.52
13.00
3.00
31931.00
0.05
0.41
8.20
3.00
32014.00
0.03
0.41
13.67
3.00
32044.00
0.05
0.81
16.20
3.00
32065.00
0.08
0.83
10.38
3.00
32254.00
0.08
0.89
11.13
3.00
32296.00
0.04
0.51
12.75
3.00
32364.00
0.03
0.51
17.00
3.00
32413.00
0.05
0.61
12.20
3.00
32440.00
0.07
0.63
9.00
3.00
32616.00
0.06
0.67
11.17
3.00
32672.00
0.03
0.41
13.67
3.00
32721.00
0.04
0.41
10.25
3.00
32765.00
0.07
0.61
8.71
3.00
32799.00
0.08
0.58
7.25
3.00
32980.00
0.06
0.68
11.33
3.00
33029.00
0.04
0.63
15.75
3.00
33087.00
0.03
0.41
13.67
3.00
33129.00
0.03
0.51
17.00
3.00
33168.00
0.09
1.11
12.33
3.00
33339.00
0.04
0.61
15.25
3.00
33457.00
0.02
0.41
20.50
3.00
33828.00
0.02
0.53
26.50
3.00
34219.00
0.03
0.37
12.33
3.00
34548.00
0.02
0.61
30.50
3.00
34900.00
0.05
0.51
10.20
3.00
35207.00
0.02
0.66
33.00
3.00
35219.00
0.02
0.43
21.50
3.00
35248.00
0.02
0.41
20.50
3.00
35299.00
0.03
0.61
20.33
3.00
35332.00
0.06
0.50
8.33
3.00
35571.00
0.01
0.44
44.00
3.00
35590.00
0.02
0.48
24.00
3.00
35612.00
0.02
0.41
20.50
3.00
35646.00
0.03
0.31
10.33
3.00
35691.00
0.02
0.31
15.50
3.00
35963.00
0.02
0.21
10.50
3.00
35983.00
0.03
0.21
7.00
3.00
36012.00
0.01
0.41
41.00
3.00
36320.00
0.03
0.45
15.00
3.00
36362.00
0.01
0.41
41.00
3.00
36376.00
0.03
0.43
14.33
4.00
30209.00
0.05
0.61
12.20
4.00
30228.00
0.09
0.73
8.11
4.00
30271.00
0.09
0.81
9.00
4.00
30305.00
0.10
1.02
10.20
4.00
30329.00
0.07
0.93
13.29
4.00
30369.00
0.06
0.87
14.50
4.00
30404.00
0.08
1.04
13.00
4.00
30434.00
0.08
0.92
11.50
4.00
30447.00
0.11
0.86
7.82
4.00
30474.00
0.06
0.54
9.00
4.00
30504.00
0.08
0.77
9.63
4.00
30531.00
0.10
0.65
6.50
4.00
30566.00
0.07
0.51
7.29
4.00
30600.00
0.07
0.76
10.86
4.00
30629.00
0.09
1.15
12.78
4.00
30664.00
0.14
1.41
10.07
4.00
30711.00
0.09
1.01
11.22
4.00
30742.00
0.16
1.06
6.63
4.00
30768.00
0.10
0.79
7.90
4.00
30791.00
0.15
0.77
5.13
4.00
30819.00
0.15
1.01
6.73
4.00
30874.00
0.14
1.02
7.29
4.00
30910.00
0.06
0.41
6.83
4.00
30943.00
0.06
0.69
11.50
4.00
30980.00
0.05
0.54
10.80
4.00
31000.00
0.10
0.99
9.90
4.00
31028.00
0.21
1.19
5.67
4.00
31089.00
0.11
1.02
9.27
4.00
31222.00
0.10
0.71
7.10
Date
Total Nitrogen (mg/L)
Total Nitrogen (Station CPF0880A)
30173.00
30209.00
30228.00
30271.00
30305.00
30329.00
30369.00
30404.00
30434.00
30447.00
30474.00
30504.00
30531.00
30566.00
30600.00
30629.00
30664.00
30711.00
30742.00
30768.00
30791.00
30819.00
30848.00
30874.00
30910.00
30943.00
30980.00
31000.00
31028.00
31089.00
31134.00
31155.00
31169.00
31211.00
31222.00
31239.00
31273.00
31320.00
31344.00
31370.00
31391.00
31428.00
31460.00
31496.00
31517.00
31551.00
31580.00
31616.00
31631.00
31680.00
31706.00
31735.00
31762.00
31784.00
31834.00
31860.00
31876.00
31909.00
31931.00
32014.00
32044.00
32065.00
32254.00
32296.00
32364.00
32413.00
32440.00
32616.00
32672.00
32721.00
32765.00
32799.00
32980.00
33029.00
33087.00
33129.00
33168.00
33339.00
33457.00
33828.00
34219.00
34548.00
34900.00
35207.00
35219.00
35248.00
35299.00
35332.00
35571.00
35590.00
35612.00
35646.00
35691.00
35963.00
35983.00
36012.00
36320.00
36362.00
36376.00
0.51
0.61
0.83
0.74
1.30
0.88
0.82
1.01
0.80
0.72
0.44
0.41
0.41
0.41
0.71
0.86
1.14
0.98
0.94
0.81
0.71
0.65
0.41
0.41
0.51
0.61
0.48
0.61
0.81
0.82
0.81
0.68
0.43
0.51
0.65
0.51
0.61
0.60
0.65
0.89
0.99
1.01
0.93
0.82
0.49
0.62
0.51
0.41
0.61
0.71
0.94
1.08
1.43
1.51
1.15
0.74
0.74
0.52
0.41
0.41
0.81
0.83
0.89
0.51
0.51
0.61
0.63
0.67
0.41
0.41
0.61
0.58
0.68
0.63
0.41
0.51
1.11
0.61
0.41
0.53
0.37
0.61
0.51
0.66
0.43
0.41
0.61
0.50
0.44
0.48
0.41
0.31
0.31
0.21
0.21
0.41
0.45
0.41
0.43

Appendix E/JOR89INT.BIN
Appendix E/JOR89INT.OUT
Appendix E/JOR90REV.BIN
Appendix E/JOR90REV.OUT
Appendix E/JOR97DET.OUT
Appendix E/JOR97INT.BAL
Appendix E/JOR97INT.BIN
Appendix E/JOR97INT.INP
Appendix E/JOR97INT.OUT
Appendix E/Read Me.docJordan Lake BATHTUB Input/Output Files

The enclosed files reflect final calibrated versions of the US Army Corps' BATHTUB eutrophication model for each of the years utilized in the scoping model analysis for the Jordan Lake
Nutrient Response Modeling Project. Reading and utilizing these files will require the installation of the Y2K compliant version of BATHTUB.

The enclosed self-extracting file, simptech.exe, contains the Y2K compliant version of the US Army Corps' Simplified Procedures for Eutrophication Assessment and Prediction, which includes
FLUX, BATHTUB, and other associated eutrophication simulation models. Documentation is also included in the self-extracting file.

Appendix E/simptech.exe
Appendix A.xlsSheet3
Tributary
In-Lake
TITLE
FILE NAME
LOCATION
SOURCE
PARAMETERS
DATES
JLDWQLA.xls
DWQ In-Lake Monitoring Data
NCDWQ
CONTACT
Debra Owen
Surface physical data, photic sone data, metals surface data
July 1982-August 1999
1999 DWQ In-Lake Monitoring Data
Jordan 99 Workbook.xls
Jay Sauber
Nutrient series, chlorophyll a, physical data parameters, solids and metals data
1999 monitoring year
Cary/Apex In-Lake Monitoring Data
JLCINTWQ.xls
Cary/Apex Water Treatment Facility
Kelvin Creech
Date, time, lake level, surface elevation, secchi depth, depth of sample, conductivity, pH, turbidity, DO, temperature, Fe, Mn
1996-1999
Jordan Physical Data
Jordan Physical Data.xls
Station, date, depth, Temperature, DO, pH, conductivity, secchi depth
2000 growing season
In-Lake Eutrophication Monitoring Data
Jleutroparms.xls
SYSTAT
monitoring data, 5,7,and 10 previous day flow, non-algal turbidity, and nonalgal turbidity to total P ratio
1982-1999
In-Lake Eutrophication Parameters
jdata.SYS
Same as Jleutroparms.xls, but in SYSTAT format
Regression trees, regressions, and correlations of In-Lake Parameters
jordan.syo and jordan2.syo
SYSTAT output for the in-lake parameters including regression trees, regressions, and correlations
Summary of Statistical Analysis
Summary of Statistical Analysis.doc
Summary of resuls and conclusions from the SYSTAT regression tree analysis
Cary/Apex In-Lake Algal Community Data
CARYALG8.xls
CARYALG9.xls
USACOE Sediment Resurvey Cross Sections
JLSED97X.txt
TJCOG
Patrick Davis
file contains elevations and coordinates for more than 16,400 points in and adjacent to Jordan Lake
Spring and Summer 1997
Select X-Sections from JLSED97X.txt
Jordan X-Section.xls
Patrick Davis/ Jason Doll
Three x-sections from sediment resurvey data
IN-LAKE DATA SOURCES
algae enumeration data for Anabaena, Synedra, and Melosira
TRIBUTARY DATA SOURCES
Haw River near Bynum Flow Data
Bynumflow.xls
USGS Water Resources of NC
Curtis Weaver
average daily flow of Haw River and several hydrographs
1972-2000
Morgan Creek near Chapel Hill Flow Data
morgancreek.xls
average daily flow of Morgan Creek and several hydrographs
New Hope Creek near Blands Flow Data
newhope.xls
average daily flow of New Hope Creek and several hydrographs
average daily flow of Northeast Creek and several hydrographs
Northeast Creek near Genlee
1995-2000
1982-2000
USGS Water Quality Data Original Data
usgsdata.db
full list of physical and chemical parameters collected and analyzed by USGS
late 1950's through 1999
USGS Monitoring Data for Specific Tributaries
usgsdat.qwp
DWQ Monitoring Data Original Data
dwqtext.db
J. Phil Bethea
full list of physical and chemical parameters collected at DWQ ambient monitoring stations
late 1960's through 1999
Bynum DWQ Monitoring
hawDWQ.qwp
June 1980-April 2000
Morgan Creek DWQ Monitoring
morold.db
Date, flow, TN, NH3, TKN, NOX, TP, OrthoP
Date, Flow, TN, NH3, TKN, NOX, TP
1974-1998
Northeast Creek DWQ Monitoring
neDWQ.qwp
Date, flow, TN, NH3, TKN, NOX, TP
January 1971-April 2000
New Hope Creek DWQ Monitoring
nhopeDWQ.qwp
January 1980-April 2000
Haw River near Bynum Monitoring Data
Bynum.xlw
USGS, DWQ, WRF Self-Mon Data
USGS and DWQ monitoring data including N, P, and flow data
June 1980-December 1998
Haw River near Bynum Nutrient Data
Hawdat.xls
Morgan Creek Nutrient Data
morgandat.xls
January 1983-October 1999
New Hope Creek Nutrient Data
newhopedat.xls
Date, Flow, TN, NH3, TKN, NOX, TP, OrthoP
December 1980-April 2000
Northeast Creek Nutrient Data
northeastdat.xls
December 1982-April 200
98 Analysis
FLUX WQ monitoring Data
FLUX analysis for the Haw River and New Hope Creek
through 1998
Haw River Near Bynum
Bynum Flux.doc
Bynum FLUX_98.doc, Durham FLUX_98.doc
FLUX analysis for the Haw River
New Hope Creek
New Hope FLUX.doc
FLUX analysis for New Hope Creek
Morgan Creek
Morgan FLUX.doc
FLUX analysis for Morgan Creek
Northeast Creek
Northeast FLUX.doc
FLUX analysis for Northeast Creek
USGS Trends
usgstrends.doc
USGS
Summary of trends for N and P in the Haw R., Morgan Cr., New Hope Cr., and Northeast Cr.
1983-1995
Comments on Jordan FLUX Analysis
jordan flux comment.doc
JASON\D:\Lauren Backup\Lauren's Work\Flux\jordan\
File summarizes all data manipulation needed for the flux analysis
OWASA Instream Self Monitoring Data
OWASA\owasa.xls
OWASA
Date, Flow/Stage, Temp, DO, pH, Cond., Fecal, TP (Conc and Mass), PO4, NH3N, TKN, NOX, TN (conc. And Mass)
September 1995-October 1999
Comments on OWASA Self Monitoring Data
OWASA\owasa_comment.doc
File contains an explanation of any data manipulations necessary to incorporate the monthly monitoring reports into file
South Durham Instream Self Monitoring Data
Durham\durham.xls
South Durham
Date, Flow/Stage, Temp, DO, BOD, pH, Cond., Fecal, TP (Conc and Mass), PO4, NH3N, TKN, NOX, TN (conc. And Mass)
Comments on South Durham Self Monitoring Data
Durham\durham_comment.doc
Facility Discharge Self Monitoring Data
JLWWTDIS.txt, JLWWWPTS.txt, JLWWTDSC.txt
first two files contain avg. daily discharges and avg. effluent conc. Data. The third file describes the records in the first two files
1989-August 1999
Depth Profile Monitoring Data
Jordan4.xls
Jordan5.xls
Surface and depth profile physical data, photic zone data, and metal surface data
July 1968-September 1997
August 1981-September 1997
M:\Jordan Lake Data\Lake Data\
M:\Jordan Lake Data\Lake Data\Jordan Physical Data
M:\Jordan Lake Data\Lake Data\Statistical Analysis
M:\Jordan Lake Data\Lake Data\Jordan Physical Data\Bathymetry Data\
M:\Jordan Lake Data\Tributary Data\USGS Data Sets
M:\Jordan Lake Data\Tributary Data\Flow Data
M:\Jordan Lake Data\Tributary Data\DWQ Data
M:\Jordan Lake Data\Tributary Data\Trib WQ Databases
M:\Jordan Lake Data\Tributary Data\FLUX Analysis
M:\Jordan Lake Data\Tributary Data\Self Monitoring
DATA COMPILATION
jdataupdated.xls
DWQ
1982 - 2000
monitoring data, 5,7,and 10 previous day flow, non-algal turbidity, and nonalgal turbidity to total P ratio; old corrected chl a and new uncorrected chlorophyll a; the corrected is good
from 1982 to Sept 96, the revised uncorrected is only available for 99 and 2000.
Usgs.xls
Water Quality Data
Usgs
web
Nutrients, organics, chl, physical, etc
1988-1999
usgspartial .xls
shows chl a data only and compares to DWQ data; they use different methods of analysis so not used for trend analysis.
Jordan 2000 Chemdata.xls
2000 DWQ In-Lake Monitoring Data
5/22/00 to 2/20/01
1_Jordan Nutchl Data 99.xls
Nutrient data, chl a (old corr and new uncorr (5/00 to 12/00)), solids, metals
Same data as Jordan 99 Workbook.xls with additional column for revised, uncorrected chlorophyll a data for 1999
Chlorodata.xls
Storet/DWQ
Algal data from Storet, corrected and uncorrected chl a for 1982 - 1997 from Storet, missing chl a for 1995 and 1998, 1999 and 2000 are revised, uncorrected chl a from DWQ, the corrected
data give for 1999 is invalid.
Chlorophyll and algal data, Fig in Data Memo report comparing uncorr and corr chl a from 1982 to 1996 comes from this file
SummerAves.xls
Uncorrected data from 1982 to 1997 is from storet, 99 and 00 from DWQ; averages are calculated from chlorodata.xls
Summer avg, uncorr chl a concentrations form 1982 to 2000, contains trend graphs for Data Memo
Alix:\C\My Documents\Jordan Lake\Chlorophyll; will be placed on file server once complete.
jordan chl jay.xls
Chl a for 1995 and 1998
Chlorophyll a data for missing years (1995 and 1998)
1995, 1998
1999-2000
Doll20010808.xls
Tributary DWQ Monitoring
Date, depth, conductivity, water temp., do, air temp, pH, precip., NH3, TKN, NOX, TP
Tributary Temp data for efdc
Tempdata.xls
Compiled
Alix Rooker
1997-2000
Jason Doll
Date, water temperature (used in tser.inp)
Appendix A A-
Appendix A A-
1998.00
1999.00

Appendix B.docJackson, J. R., Rice, R. A., Nobel, R. L., and Mozley, S. C. Mechanisms of Reservoir Fish Community Dynamics. [Jordan Lake]. Federal Aid in Fish Restoration Project
F-30-1. North Carolina Wildlife Resource Commission, Division of Boating and Inland Fisheries. 1992. B-4
Kuenzler, E.J., Belense, A. J., and Rudek, J. Nutrient Cycling and Productivity of a North Carolina Piedmont Reservoir. WWRI Report No. 228. Water Resources Research Institute of
the University of North Carolina, Raleigh, N.C. 1986. B-7
Weiss, C. M., Francisco, D. E., and Campbell, P. H. Water Quality Study, B. Everett Jordan Lake, North Carolina Year II, December 1982-November 1983. A Report to the Wilmington District
of the U.S. Army Corps of Engineers. 140 pp., July 1985. B-16
Weiss, C. M., Francisco, D. E., and Campbell, P. H. Water Quality Study, B. Everett Jordan Lake, North Carolina Year III, December 1983-November 1984. A Report to the Wilmington District
of the U.S. Army Corps of Engineers. 120 pp., October 1986. B-18
Smith, V. H. Prediction of Nuisance Blue-green Algal Growth in North Carolina Waters. WWRI Report No. 233. Water Resources Research Institute of the University of North Carolina,
Raleigh, N.C. 1987. B-19
North Carolina Division of Water Quality. Water Quality Conditions B. Everett Jordan Reservoir 1996-1997. 16 March 1999. B-21
Hoyer, M.V. and Jones, J. R. “Factors Affecting the Relation Between Phosphorus and Chlorophyll a in Midwestern Reservoirs.” Canadian Journal of Fisheries and Aquatic Sciences. 40
(2): 192-199. 1983. B-23

Jackson, J. R., Rice, R. A., Nobel, R. L., and Mozley, S. C. Mechanisms of Reservoir Fish Community Dynamics. [Jordan Lake]. Federal Aid in Fish Restoration Project F-30-1. North
Carolina Wildlife Resource Commission, Division of Boating and Inland Fisheries. 1992.

Phytoplankton and Depth
“Secchi disc values during the course of the study were highly variable, in part due to sediment transport associated with high periods if high water inflow and the cycle of resuspension
and settling of clay particles caused by wind generated turbulence” (10)
“Seasonal trends in phytoplankton density indicate that the drop in Secchi depths observed in the summer was in part due to peaks in plankton production” (10)
Weiss reported linear trend in Secchi depth in New Hope arm that was consistent with phytoplankton distribution. (The upper New Hope arm had lower Secchi depths than the middle arm
and the lower arm had a slightly greater transparency than the middle arm.) (10)

Nutrients
Comparisons of inflow and outflows illustrate “[a]s much as 86 percent of total phosphorus is lost from streams flowing into the New Hope arm, while 48 percent of total nitrogen is lost.”
This results in a pronounced nutrient gradient down the main body of the lake (10-11)
“Nutrient sedimentation down the New Hope arm is facilitated by the limited water flow between basins caused by highway causeways …” (11)
Nitrogen and phosphorus concentrations are highest in winter and “lowest in early and late summer when assimilation by phytoplankton peaks” (11)

Reservoir Community Dynamics: Relationship of Nutrients and Primary Productivity (20-21)
“Weiss, et al. (1985) employed the Dillon-Rigler model, which relates total phosphorus inputs to phytoplankton production, as measured by chlorophyll a, and found predictions to consistently
overestimate algal production in the New Hope arm.” They concluded total phosphorus was only one of several possible factors contributing to algal growth (20)
“Kuenzler et al. (1986) found that neither phosphorus nor nitrogen alone control algal production in Jordan Lake, and pointed out the importance of temperature and light penetration”
(20)
“Weiss, et al. (1986) … concluded that a useful model of Jordan Lake’s nutrient/phytoplankton relationships would have to incorporate hydrodynamic, climatological, and morphological
variables” (20)
“A review of the history of efforts to model Jordan Lake’s primary productivity reveals that the unique hydrological patterns and sensitivity of water movement to localized events may
confound efforts to create a predictive model. The restricted exchange sites between the basins of the New Hope arm, combined with extreme sensitivity of both flow direction and retention
time to inflows results in variable rates of nutrient transport, assimilation, and settling. High levels of turbidity, combined with frequent wind mixing, result in mixing depths which
are frequently deeper than the depth of light penetration. As a result, at certain times of the year, light availability may limit phytoplankton production even when sufficient nutrients
are available” (21)

Eutrophication, Algal Blooms
NC Tropic State Index (NCTSI); Jordan Lake – 3.9 in 1987, alpha-eutrophic (11)
“Jordan Lake has not demonstrated a tendency to support noxious algal blooms. The rapid loss of nutrients from inflowing streams, combined with the short retention times associated
with high water and nutrient flows, favors small celled algae with rapid growth rates …” (12)

Graphs of Possible Interest
Typical annual temperature profiles from a deep water station (67), shallow station (68)

Phytoplankton
Weiss et al. collected monitoring data from 1984-1988. They found that “[m]ean annual values for chlorophyll a, density, and biovolume indicate that plankton production reflects nutrient
availability in its downstream distribution.” Values for middle New Hope arm (Basin III) were generally only 65-70 percent of those of the upper New Hope arm (Basin IV); lower New
Hope arm values (Basin II) ranged from 55-60 percent of the Basin IV values (11)
The three most important types of phytoplankton in Jordan Lake are blue-green algae (dominate in biovolume and density), the diatoms (important in spring) and the green algae (dominance
following impoundment; important in winter) (11)

Plankton Productivity Model
“We constructed a model to determine if spatial and temporal variations in primary production could be related to fish production. … Productivity relationships among the three basins
of the New Hope Arm were modeled for the years 1982-1985 using the data of Weiss et al. (1984, 1985, 1986, 1988) and for the years 1983-1988 using North Carolina Department of Environmental
Management (NC DEM) data. ” The model’s predictions showed limited variation between years (20)
“No significant basin differences were predicted by the model. This result is inconsistent with distributions of nutrients and phytoplankton reported by Weiss et al. (1988), and may
indicate that gross primary productivity is not exclusively determined by nutrient levels and phytoplankton biomass. Physical factors such as turbidity and wind generated mixing may
operate to dampen the differences in productivity that would normally be associated with nutrient concentration” (20-21)

Appendix 1: Plankton Productivity Modeling
Jackson et al.’s objective was “to adapt and calibrate existing models of phytoplankton primary production for estimation of the Jordan Reservoir food chain basis in relation to space
and time” (1)
The data were collected from fall 1983 until mid-1985 and consisted of primary productivity and respiration measurements, nutrient concentrations, temperature, transparency, and chlorophyll
concentrations (1)
An equation was constructed with the available data and fit to observed gross primary production (GPP) measurements. “Then, data on the independent variables from the two long-term
monitoring projects (Weiss et al. and NC DEM) were substituted into model to estimate GPP…” (1)
The variables entered into model were temperature squared (T), square root of chlorophyll (Chl), light attenuation (compensation depth, Zc), solar irradiance (maximum daily potential
solar flux, IF) nitrogen effect (120/ dissolved inorganic combined nitrogen (DIN)= NF; but if 120/DIN<1, NF=1 ), phosphorus effect (total phosphorus (TP)/50=PF, but if TP/50<1, PF=1)
(2-4)
Final Model: GPP = .332 * Ln (T2 * squareroot (Chl) * Zc * IF * NF) – 1.1
The model was primarily influenced by temperature (4)
“The variation in total GPP estimates among basins and years was minor.” This result is consistent with the direct GPP measurements by Kuenzler et al. (1986) in the upper and middle
New Hope arms, which show no major differences among basins (5)
“Comparison of GPP estimates on different data sets collected in the same year give a rough idea of the potential error in this procedure. … In view of the differences among estimates
from different data sets for almost the same place and time, it would be inappropriate to consider the minor differences among estimates for years and basins as significant.” (5)
“One is forced to the conclusion that regulation of GPP in Jordan Lake either has nothing to do with nutrients, light or phytoplankton biomass, or that there are offsetting factors which
level the rates between basins” (e.g., nutrient loading and clay turbidity) (6)

Kuenzler, E.J., Belense, A. J., and Rudek, J. Nutrient Cycling and Productivity of a North Carolina Piedmont Reservoir. WWRI Report No. 228. Water Resources Research Institute of
the University of North Carolina, Raleigh, N.C. 1986.

Sedimentation
Jordan Lake’s “Piedmont watershed supplies abundant suspended sediments, causing lake water to be turbid with suspended clays” (xiii)

Physical and Chemical Factors Affecting Algal Biomass and Productivity
Seasonal patterns suggest changing light and temperature may be controlling factors of algal abundance and primary productivity (xiii)
“Although algae abundance was poorly correlated with temperature, gross productivity showed strong correlation” (xiii)
Light extinction did not show clear seasonal pattern and was not correlated with temperature (xiv)
River discharges and seasonal change have major effects on Jordan Lake’s water quality (75-77)
“Heating not only reduces viscosity, thereby increasing sinking rates of seston (suspended materials, including phytoplankton) but also causes water column stability, reducing the vertical
turbulence which would return seston to the surface (Wetzel 1993).” (75)
“Temperature of Jordan Lake surface waters, however, did not correlate significantly with light extinction, phosphorus concentrations, or ammonium concentrations, but showed a strong
negative correlation with nitrate” (75)
“There was clearly a correlation between Haw River and New Hope River discharges 5 days prior to sampling and light extinction at Stations 5 and 30, respectively” (77)

Phytoplankton
“The major factors controlling phytoplankton growth in nature are usually temperature, light, or one of a few nutrient elements.” (2)

Chlorophyll a
Chlorophyll a had an overall three station mean from April 1984-January 1985 of 40 +/- 36 g/l. The concentrations were highest in August and September but exceedences of the standard
(40 g/l) also occurred in samples collected in July, November, December, and January (36)

Inflows
Discharges from Haw River, Morgan Creek, New Hope Creek, and Northeast Creek each varied more than two orders of magnitude, resulting in pulse inputs of water, nutrients, and suspended
sediments (16)

Productivity
Brylinsky and Mann attempted to evaluate factors controlling primary productivity rates “using data collected during the International Biological Program. Multiple regression analysis
of the IBP data suggested that variables related to solar energy input were more important than variables related to nutrients.” (14)
Weiss and Kuenzler (1976) measured primary productivity in 9 lakes within 125 miles of Jordan Lake … productivity ranged from 3-188 mg C/m3 h (15)
Measurements of gross productivity in Jordan Lake showed “low gross and net productivities of the cool seasons and high rates from June through September.” “Highest gross productivities
were also found at Station 30 [Haw River arm] in July and August, with a maximum of 1.4 mg O2 / l * h. Station 10 [middle New Hope arm] generally showed the lowest productivity and
respiration rates.” (70)
“The seasonal changes and station-to-station differences among gross productivities are also apparent when the rates at each measured depth were integrated down through the euphotic
zone, corrected for day length, and converted to carbon units. Station 30 [Haw River arm] dominated the other two stations during June – August when rates exceeded 4 g C/ m3 d [167
mg C/ m3 h] while the other stations showed rates < 3 g C/ m3 d [125 mg C/ m3 h]. Rates at all stations in February, April and December were < 1 g C/ m3 d [42 mg C/ m3 h].” When “the
24 hour respiration of the whole water column is subtracted from the gross production of the photic zone to obtain net water column productivity”, “[o]nly Station 30 in August, Station
10 in October and all three stations in September had net water productivities >2 g C/ m3 d [83 mg C/ m3 h].” (70)
“The seasonal variation in gross productivity and ln gross productivity was strongly correlated with temperature” (99)
“Photosynthesis was typically inhibited by excessive light intensity in the bottles incubated closest to the surface; the rate was highest at somewhat greater depths, and then decreased
approximately exponentially as depth continued to increase.” Respiration also tended to decrease with depth (103)
There was a seasonal pattern of photosynthesis and respiration; the rates were low in February and April 1984 and progressed to high rates in June through September (103)
“Linear correlations demonstrated strong positive relationship of gross productivity (g O2 /m2 d) with temperature and negative relationships with light extinction and NO3 concentration.
Turbid waters with high light extinction coefficients caused decreased photosynthetic rates. Periods of high productivity depleted nitrate in the water.” (104)
Transformation to ln gross productivity resulted in signification correlations with temperature, light extinction, filterable reactive phosphorus, NO3, NH4, and particulate nitrogen
(see Table 13 on page 48 for coefficients) (104)

Tropic State Index
“Weiss and Francisco (1985), using their 1982 and 1983 data and the trophic state index … recognized spatial variability in Jordan Lake. The New Hope River segment tended to be meso-
to alpha-eutrophic in early-summer …” (105)
“The equations of Carlson (1977) were used to examine spatial and seasonal variation in trophic state index (TSI) in Jordan Lake. The general pattern of changes in TSI were the same
at all stations.” The findings are based on data from February 1984 to January 1985 and are illustrated in Figure 26 on page 107. “The TSI based on chlorophyll-a increased from April
to July or August and then decreased again at all stations.” The TSI based on total phosphorus and the TSI based on Secchi depth “decreased from February to lower, irregular values
in May.” The station on the middle New Hope arm showed some trends that differed from the other stations including a second increase in the chlorophyll-a TSI from November to January
and a drop in the total phosphorus and Secchi depth TSI values after November (106)
“[R]egardless of the base chosen, Carlson’s TSI equations place almost all stations in eutrophic categories almost all of the time.” (106)

Studies relating environmental factors to phytoplankton nutrition and growth
Weiss and Kuenzler (1976) “examined more than 20 physical, chemical, and biological characteristics of 69 lakes, reservoirs, subsegments of reservoirs and river segments” and developed
scales of trophic status (3)
“Weiss (1976) conducted 345 algal assays experiments in surface waters from many parts of North Carolina to determine nutrient limitations in lakes, rivers, and impoundments” (3)
“He related the results of his assays to the chemically measured concentrations of N and P, and to the N/P ratios. P-limited waters had N/P ratios of >16 (by weight), N-limited waters
gad ratios of 5-7 and N- and P-limited waters had ratios of 9-11. He judged that an increase in algal abundance of >5.1 mg dry wt./l after addition of a nutrient demonstrated a significant
response to that nutrient.” (13)
Weiss showed that “11 stations on 9 Piedmont impoundments within 125 km of Jordan Lake were all P limited rather than N limited” (14)
Kuenzler and Greer (1980) “showed that phytoplankton is not the only agent removing phosphorus from the water; for most of the year ‘bacterial’ uptake and uptake by suspended sediments
was very important.” (3)
Studies on phytoplankton nutrition and growth in Chowan River, N.C. made by Stanley and Hobbie (1977), Sauer and Kuenzler (1981), Kuenzler et al. (1982), Peaerl (1982) … “demonstrated
that either N or P may at times limit algal growth, but that species composition and algal abundance often are controlled by other factors such as light, temperature, salinity, or humic
substances.” (3-4)
Sauer and Kuenzler (1981) “showed that when N and P are simultaneously in short supply, large increases in one or the other gave only a slight increase in algal growth” (4)
“No reports have been seen which attribute algal nutrient limitation in North Carolina to an element other than N or P” (14)

Nutrients
“Correlations with nutrient concentrations were negative, indicating that algal growth was controlling nutrient concentrations rather than nutrients controlling algae” (xvi)
“Our study showed high total N and total P concentrations in lake water during most of the year.” (xvii)
Because nitrogen-fixing cyanobacteria are least desirable and thrive in P-rich waters, “control of P inputs will probably be more effective than control of N inputs” (xvii)
“Because clays appear to be carrying phosphate to the bottom of the lake, research seems appropriate to determine the capacity of bottom sediments for long-term storage of P.” (xviii)
P, which is generally considered to limit algal growth in freshwater systems, may not “when nutrient inputs have a low N to P ratio” (4)
“The Redfield ratio (Redfield 1958) predicts N:P atom ratios greater than 16 can produce P limitations while ratios less than 16 can produce N limitation” (4)
“The algal assay technique seldom indicated severe limitation of phytoplankton growth by N alone or P alone” (68)

Phosphorus Cycling in Lake Water (5-11)
“During periods of phosphate deficiency phytoplankton and bacteria can develop phosphatase enzymes that allow for the utilization of dissolved organic phosphorus compounds which are
analogous to the XP fraction (Kuenzler and Perras 1965; Paerl and Downes 1978).” (5)
“Many phytoplankton species can take up P in excess of their metabolic needs and store it within the cell as polyphosphate (Perry 1976)” Algae can use this internal source when P is
limited in water (5-6)
“Nutrient dynamics in lake waters cannot be understood simply by measuring concentrations. Periods of maximum biotic activity often correspond to periods of low P concentrations that
should, if recycling is ignored, limit further biological activity.” This indicates the importance of measuring nutrient flux. (6)
“The rate of P uptake by micro-organisms is affected by many factors including: nutritional status, presence or absence of light, and light intensity; pH; cell size; temperature; diurnal
cycle; and phosphorus concentration” (report includes references for each factor) (7)
“Suspended sediments carried into lakes may be important to the P cycle (Hutchinson 1941, Kuenzler and Greer 1980, Jones and Redfield 1984; Cuker 1986), especially when the sediments
are fine clays and have high P binding capacities and long retention times in the water column (Golterman 1973; Syers et al. 1973)” (7)

P desorption/adsorption capacity of suspended sediments (8-11)
“While most studies have found that sediments act as a sink for phosphate (Syers et al. 1973), the complexity of the sorption reaction combined with the large variability found within
and between aquatic systems makes generalizations difficult” (8)
“The sorption mechanism is a function primarily of the solution phosphate concentration and pH (Chen et al. 1973a), but ionic strength, substances competing for the same reactive sites,
temperature, organic matter, retention time, and particulate size are also involved (Beek and vanRiemsdijk 1982)” (8)
Reactions between phosphate ions and clay mineral surfaces are not well understood. These reactions are associated with the presence of aluminum hydroxides and iron oxides. Studies
have shown the “uptake of phosphate by clay to be biphasic, consisting of a rapid initial adsorption followed by a much slower phase which may involve the formation of new solid phases
…” (8-9)
“Quantification of the amount of P bound to suspended sediments that is available for algal utilization may help explain the role of these sediments in the P cycle of lakes.” (9)

Bioavailability of sediment-bound Phosphorus
available for algal utilization may help explain the role of these sediments in the P cycle of lakes.” (9)
“Laboratory experiments attempting to quantify the amount of sediment P available to algae have primarily used modifications of standard bioassay methods.” (10)
“The availability of P to algae under natural conditions is a function of many factors including: the forms and amounts of P in the particulate fraction, the residence time of the particle
in the lake water, the abundance, species composition, and nutrient status of the algal population, the solution phosphate concentration, and other factors controlling particulate-P
solubility, such as pH and Eh (Armstrong et al. 1979)” (10)

Nitrogen Cycling in Lake Water (11-13)
Dugdale and Goering (1967) “proposed a primary production model based on a simplified N cycle. In this model, newly available N enters the system as nitrate. Phytoplankton assimilation
of nitrate is therefore associated with ‘new production’ …” (11)
“Ammonia is made available as a nutrient to the phytoplankton by regeneration from organic matter, by zooplankton excretion, bacterial remineralization, or leakage from photoplankton
cells themselves (Brezonik 1972). Primary production associated with ammonium assimilation is therefore termed ‘regenerated production’” (12)
Ammonia is favored as N source for phytoplankton since it does not require reduction as nitrate does (12)
The following studies on regenerated N on lake systems “indicate that ammonia assimilation is very important and may even control primary production, especially during algal blooms”:
Alexander (1970), Brezonik (1972), Toetz and Cole (1980), Axler et al. (1981) (12)
“Ambient dissolved inorganic N concentrations can be very low when a dynamic balance exists between ammonium removal and regeneration” (12)
“Nitrogen uptake measurements are a function of incubation time during the tracer study, physiological conditions of the cell, light levels, and temperature as well as N concentrations.”
(13)
“One of the largest problems in calculating N-uptake rates is the inability to measure very low concentrations of N accurately. Goldman and Gilbert (1983) provide a comprehensive review
of the kinetics of nitrogen uptake” (13)

Phosphorus Distributions, Uptake Kinetics, Sorption, and Bioavailability (38-54)
“The distributions of phosphorus fractions in Jordan Lake appeared to be controlled largely by hydrologic events but were also affected by biological processes.” (38)
“Total phosphorus (TP) averaged 120 g/l, clearly indicating the enriched status of the lake. Spatial variability of phosphorus was very pronounced. “Station 30 [Haw River arm] always
had the highest TP concentrations (198 (141 g/l), followed by Station 5 [upper New Hope arm] (118 (93 g/l), and then by Station 10 [middle New Hope arm] (69 (50 g/l), showing the
trend of decreasing TP with distance from the ends of the reservoir.” (38)
“The relative rate of FRP [filterable reactive phosphate] uptake (K) tended to be low during winter at all three stations and high during the warmer months at Stations 5 and 10” (41)

Total gross uptake rates “were lowest and most nearly constant at Station 30 where TP concentrations were highest. Rates at Stations 5 and 10 had significantly more variation with time
and tended to be higher during the warmer months.” (41)
“Biotic uptake was dominated by the small size fraction [0.45 to 8.0 m fraction, initially assumed to be mostly bacterial], averaging 87.6 ( 7.9% of total biotic uptake” (41)
“Uptake rates down the water column in March at Station 10 decreased nearly threefold, from 24.8 g/l h at 0.2 m to 8.3 g/l h at 4 m” (45)
“About 90% of the biotic uptake was insensitive to antibiotics in March and 58% in May, suggesting that most uptake was procaryotic” (48)
“The effect of additional phosphate substrate on uptake rates was measured in August 1984 and May 1985 at Station 10.” Results indicated “the rates of biotic uptake were limited by
the available P.” (48)
“High TP concentrations in the Haw River presumably were due to the high P content of suspended clays in combination with the numerous agricultural and industrial sources of P in the
Haw River watershed” (49)
Collection of data on successive days demonstrate “the large errors that may occur when estimates of the net effect of suspended sediments are made using data from only one sampling
during high flow conditions” (53)
“The algal availability P (AAP) associated with suspended sediments ranged from 5.1 to 18.6% (mean =11.3%) of the total sediment” (54)
Experiments showed NaOH is a very poor indicator of AAP (54)

Nitrogen Distributions and Uptake Kinetics (54-68)
“Ammonium concentrations declined in both arms of the lake from February 1984 to August 1984, with levels at or near the detection limits during the summer” (57-58)
“Nitrate levels in the New Hope arm were very high during the cold months but decreased to levels below detection limits during the warmer months. The Haw River station had nitrate
levels below detection limits only during the months of July and August 1984; otherwise levels were very high in this arm” (58)
“Nitrate was more abundant than ammonium in Jordan Lake whenever these nutrients were detectable” (58)
“Particulate N concentrations were generally high at all stations” (58)
“The largest fraction of nitrogen present in the epilimnion was usually the dissolved organic N” (58)
“Ammonium uptake rates in the Haw River arm steadily increased from May until October when rates peaked at 26 g / l * h.” The New Hope arm also had high rates during this period.
“Nitrate uptake was also rapid, but generally slower than ammonium uptake.” (61)
Ammonium and nitrate turnover times in the winter “were on the order of days or weeks, whereas during the warm months they were only about a half an hour to a few hours.” (61)

Algal Biomass in Relation to Nutrient Concentrations
While chlorophyll a was positively correlated with particulate nitrogen (PN) (r=.585 at p(.001), both chlorophyll a and PN were negatively correlated with filterable reactive phosphorus
(FRP) (r=-.374 at p(.05 and r=-.470 at p(.01, respectively) and chlorophyll was negatively correlated with nitrate (r=-.390 at p(.05). The negative correlations indicate the “effective
removal of nutrients during times of phytoplankton abundance” (Kuenzler 77-79)
Neither chlorophyll nor PN was significantly correlated with water temperature, light extinction, or ammonia concentration (77)
The “lack of correlation of chlorophyll and PN with particulate phosphorus (PP) is consistent with the explanation that, after heavy river discharge in winter, a major portion of PP
is suspended whereas during summer months a major portion is phytoplankton” (79)

Phosphorus Distributions and Algal Biomass
“There were clear correlations (P< 0.01) between stream discharges five days before each sampling trip (Hill et al. 1984; USGS data file) and FRP concentrations at Stations 5 and 10”
(upper and middle New Hope arms) (79)
“there were strong positive correlations (P<=.001) of PP with 5-day prior discharge at Station 5 and with 10-day prior discharge at Station 30” (Haw River arm) (79)
“FRP concentrations were strongly correlated with flow at Station 5, and PP and TP were correlated with flow at each end of the lake. Francisco (personal communication) suggested that
it usually takes more than 5 days for stream discharges to affect these stations and that these higher P concentrations in the water might originate from wind mixing and bottom sediment
disturbance” (80)

FRP Uptake Rates and Turnover Times (80-83)
“Phosphate turnover times in Jordan Lake were often short in summer and early winter at Stations 5 and 10 [upper and middle New Hope arm] when FRP was scarce and N:P ratios were high;
this also implies P limitation.” (80)
“Long turnover times and high FRP concentrations at Station 30 [Haw River arm], however, indicate that P-limitation was unlikely in this section of the lake.” (80)

Algal versus bacterial uptake of P (83-85)
“Small particle uptake was well correlated with large particle uptake at Station 5 (r=0.88) and Station 10 (r=0.73), suggesting that algal and bacterial uptake increased or decreased
together and that algae did not dominate uptake during periods of high biomass and P-deficiency.” (83)
“Analysis of the antibiotic-treated samples showed that moderate-to-rapid treatment often remained after antibiotic treatment, indicating either the presence of small eucaryotic algae
there and/or incomplete suppression of procaryotic P uptake.” (84) Algae “small enough to pass through a 8.0  filter are abundant in Jordan Lake (Weiss and Francisco 1985).” (84)

Suspended Sediment Phosphorus Sorption Capacity (85-88)
75% of total-P load in 81-82 came from Haw River, most during high flows (85)
“suspended sediments were a sink for phosphate in river water during these periods of high flow.” (85-86)
“Algae used from 5.1 to 18.6% of sediment-bound P despite indications from the sorption experiments that the sediments were undersaturated with respect to phosphate.” (88)

Management Implications of Sediment Sorption and Bioavailability
“It appears then, that suspended sediments are not the ultimate source of easily desorbable and available P, but are usually a sink for P during times when P is abundant. High ambient
FRP concentrations, and often low biomass, immediately after high river flows decrease the importance of sediment-bound P. However, as algal biomass and photic zone depth increase
and dissolved nutrient supplies decrease, the bioavailability of P on the clay particles remaining in suspension becomes more important.” (89)
“In spite of several mechanisms by which sediment-associated P can be returned to the water and to the euphotic zone, large amounts of P are lost each year to the lake bottom. Weiss
and Francisco (1985) calculated that 45-55% of TP [total phosphorus] transported by the Haw River and 84-92% of TP transported by the New Hope River were lost in the lake.” (91)
Denitrification led to the loss of 2-15% of total nitrogen (TN) from the Haw River and 57-66% of TN from the New Hope River being lost in the lake (91)
“… the unusual turbidity in our lakes may keep their trophic status below that predicted from P loading alone (Pearce 1983).” (91)

Nitrogen Distributions and Time Course of Nitrogen Uptake
“correlations between NO3, NH4, or total N concentrations and discharge were not significant” (92)
“The increase in total N concentrations in autumn was due primarily to the increase in nitrate while particulate N decreased and DON stayed the same.” (92)
“Understanding nitrogen tracer experiments in relation to primary productivity requires recognition that uptake and growth processes are not tightly coupled. A portion of the DIN taken
up by phytoplankton may be subsequently excreted, leaked, or stored by the cell (Wheeler 1983). … However, after ln transformations, we found that NH4-uptake rates were clearly correlated
(r=0.566**) with gross productivity whereas NO3-uptake rates were not” (93)

Relative Preference Indices for Ammonium and Nitrate Uptake
A log-log graph of the Relative Preference Indices “shows algal preference is predominantly for NH4 over NO3 at all stations.” (98)

Vertical Profiles of Nutrient Uptake and Productivity
“… gross productivity, and hence photosynthetic activity peaked at 1 m while NO3 uptake peaked at .5 m” (98)

Nitrogen to Phosphorus Ratios and Nutrient Control of Phytoplankton Growth
DIN:DIP ratios “suggest that algal growth is frequently limited by P” (100)
“Algal assays conducted on water from Jordan Lake showed that neither N nor P alone were consistently at concentrations low enough to limit further phytoplankton growth” … “Therefore
it appears that reduction of either nutrient … will contribute to control of algal growth (101-103)

Weiss, C. M., Francisco, D. E., and Campbell, P. H. Water Quality Study, B. Everett Jordan Lake, North Carolina Year II, December 1982-November 1983. A Report to the Wilmington District
of the U.S. Army Corps of Engineers. 140 pp., July 1985.

Provides a summary of water quality and quantity conditions in Jordan Lake for the second year after impoundment.

Phytoplankton
“Phytoplankton abundances in B. Everett Jordan Lake appear to be in part limited by factors such as temperature and low light levels, particularly in the cooler months” (xvii)

Chlorophyll a
“Our attempt to use the Dillon-Rigler model to predict chlorophyll a for B. Everett Jordan Lake using data from Years I and II, identified a variation of five orders of magnitude for
the same TP concentration. This suggests that TP at least is not the only factor controlling chlorophyll a.” (xviii)
It appears that the retention times of both arms is a major factor in establishing net chlorophyll values in relationship to the quantities of nutrients available and the manner in which
these are recycled.” (xviii)
“Pearse (1983) has evaluated the applicability of eight models which predict the chlorophyll a concentration as a function of TP or TP and TN to 40 South Carolina reservoirs. He concluded
that none of the models he evaluated, which were developed from other data bases, adequately modeled chlorophyll a in South Carolina reservoirs” (117)
Chlorophyll models “are limited in their predictive capability outside of their own data base” (117)

Seasonal and Hydrologic factors (Chapter 2)
Five seasons defined for annual data sets – winter, spring, early summer, late summer, and fall (6)
Permanent filling of the lake began September 1, 1981 (9)
Retention time can vary from 2-3 days to more than 1000 days (12)

Nutrient Relationships (Chapter 4)
Models reflect load as a function of streamflow (34)
DEM automatic sampler data from Haw indicates “a large portion of the total annual loads enters the reservoir in a relatively few storm events” (34)
“In the New Hope arm euphotic zone TP concentrations suggest that more than 90% is lost from the inputs. The data suggests, therefore, that the euphotic zone TP concentration is largely
determined by lentic processes (sedimentation / mixing) rather than the input” (52)
“The TN:TP values suggested mixed limitation” (52)
“OP was quite low or below detection limits during much of the growing season in the New Hope arm suggesting phosphorus recycling” (52)
Inorganic nitrogen also low during growing season. “The 1983 increase in inorganic nitrogen in late summer and fall was associated with mixing events. These were accompanied by large
increases in chlorophyll a” (52)

Year to Year Comparison of the Trophic State (Chapter 5)
“It is assumed that temperature and light will, from year to year, remain relatively constant, within the limits expected for this geographic location.” Nutrient input depends on streamflow,
but will remain within a limited range on an annual basis. “Therefore, a basis for year to year comparison would be to identify the direct and indirect measures of phytoplankton numbers
and metabolism, that lend themselves to calculation of a mean value over twelve months of sampling” (53)
“Whole-lake annual mean values for any biological dimension in B. Everett Jordan Lake are rendered meaningless because segmentation of the lake by road causeways affects flows through
and gradients are quickly established within segments.” Comparing annual means at sampling individual sampling locations appears to be a reasonable procedure (53)
DEM Trophic State Index calculations show the net trophic classification is mesotrophic (60-62)

Chlorophyll models (Chapter 9)
Application of Dillon-Rigler model to Jordan Lake; it is not very useful (117-119)
“The fact that chlorophyll a varies by as much as 5-fold for the same TP concentration suggests that TP is not the only factor controlling phytoplankton growth” (119)

Weiss, C. M., Francisco, D. E., and Campbell, P. H. Water Quality Study, B. Everett Jordan Lake, North Carolina Year III, December 1983-November 1984. A Report to the Wilmington District
of the U.S. Army Corps of Engineers. 120 pp., October 1986.

Nutrient Relationships (Chapter 3)
On Haw Basin, a two-day event on February 14 and 15, 1984 “accounted for 10% of the estimated annual TP load and transported the equivalent of 61 days of point-source-derived TP” (30)
“The storm load of phosphorus is largely particulate. This rapidly settles from the water column into the deeper waters. Since the storm flow is released from the bottom gates of the
discharge structure, much of the storm-derived phosphorus is almost immediately discharged. This phosphorus, a significant proportion of the annual load, has little impact on the water
column phosphorus concentration.” (30)
“During winter and early spring higher rainfall, complete mixing and decreased biological activity combined to result in a high mass of both TN and TP.” TN has large NOx component in
winter (32)
Total mass of TP determined primarily by storms but total mass of TN less affected by storms (32)
“there is a large loss of TP and a lesser loss of TN to the bottom of the lake very near the inflowing streams” (35)
Components of TN (TIN, PON, DON) do not change uniformly with TN throughout the year (39)
Low values for OP and high biomass imply P is recycled rapidly (39)
The low concentrations of biologically available forms of N and P suggest that recycling of one or both nutrients in necessary to maintain phytoplankton growth (43)

Chlorophyll Models (Chapter 6)
“[W]e evaluated correlation between chlorophyll a, both log-transformed and untransformed and a variety of water quality variables (NH3, TIN, OP, TDP, TP, phytoplankton biovolume, Z1%,
water temperatures, and these variables log-transformed.) The results showed chlorophyll a was positively correlated with biovolume and water temperature and negatively correlated
with Z1% and TIN. The other correlations were either not significant or inconsistent.” (101)
“If there is a useful model, it must include hydrodyanmic, climatological, and morphological variables as well” (101)

Smith, V. H. Prediction of Nuisance Blue-green Algal Growth in North Carolina Waters. WWRI Report No. 233. Water Resources Research Institute of the University of North Carolina,
Raleigh, N.C. 1987.

Factors influencing Algal Biomass
Studies of lakes and midwestern reservoirs and “the results of this study suggest that total phosphorus is most likely the primary limiting factor for total algal biomass in turbid North
Carolina reservoirs. However the same is not true for the summer mean biomass of blue-green algae” (vii)
Further studies on blue-green algal growth should include “detailed experimental examination of the roles of inorganic suspended solids on nutrient availability, light extinction, and
algal sedimentation” as well as the roles of hydraulic flushing and inorganic carbon availability (vii-viii)

Algal Biomass Models
“Test of 4 published empirical models for blue-green algal biomass (Smith 1985) and a model of blue-green algal relative to biomass (Smith 1986) clearly indicate that the response of
blue-green algal growth in North Carolina reservoirs to nutrients and light availability differs from that observed in clear north temperate lakes. Furthermore, stepwise regression
analysis suggests that blue-green algal biomass in North Carolina reservoirs is much more strongly influenced by non-nutrient factors such as station depth, water column mixed depth,
and non-algal turbidity.” (vii)
Smith developed models predict “the summer mean biomass of blue-green algae in northern temperate lakes from total phosphorus, total nitrogen, and lake mean depth.” He also developed
a model that predicts “the summer relative biomass of blue-green algae (%BG, the proportion of the total algal biomass which is represented by blue-greens)” from Secchi depth and summer
mean mixed depth. “Recent research has in fact suggested that models developed from north temperate lakes may not apply to the warmer, more turbid water resources of the southeastern
United States.” (2-3)
“The dummy variable analysis indicated no significant difference in the response of total algal biomass to phosphorus in clear lakes and North Carolina reservoirs.” However there was
a significant difference in blue-green algal biomass (5)

Non-Algal Turbidity
“In order to evaluate the possible role of inorganic turbidity, and estimate of non-algal turbidity was calculated for each station from measurements of Secchi disk transparency. …
A plot was first made of Secchi disk transparency versus measured concentrations of chlorophyll a, and then the values of A and B in the generalized relationship 1/SD = A + B(CHL) were
selected to provide good empirical fits to the upper and lower bounds of the data.” In the equation above, “the value of A is an estimate of non-algal turbidity, and B is an estimate
of the extinction due to chlorophyll a.” (10, 17, 19).
“non-algal turbidity should be measured directly as concentrations of inorganic suspended solids” (26)

Blue-Green Algae
“Simple regression analysis further indicated that the summer mean blue-green biomass in these reservoirs was only very weakly correlated with total phosphorus” (20)
“slow-growing blue-green algae may be held in check by the rapid flushing rates observed in lakes and reservoirs with extensive fluvial inputs” (Uttormark and Hutchins 1985) (22)
Low inorganic carbon availability has been “experimentally demonstrated to favor dominance by blue-green algae” (23)
“Perhaps the most likely factor influencing the growth of blue-green algae in North Carolina reservoirs is the presence of high concentrations of inorganic suspended solids” … “high
concentrations of inorganic suspended solids clearly can reduce the concentration of chlorophyll a produced at a given concentration of total phosphorus" (23-24)
“presence of high levels of inorganic turbidity acted to reduce the likelihood of blue-green blooms in eutrophic reservoirs” (24)
The mechanisms of silts and clays that thwart growth of blue-green algae are P binds to sediments, sedimentation, and their effects on light penetration and scattering (24-26)

North Carolina Division of Water Quality. Water Quality Conditions B. Everett Jordan Reservoir 1996-1997. 16 March 1999.

Hydrologic factors
“The Haw River arm of the lake has an average hydraulic retention time of five days an accounts for 70-90% of the annual flow through Jordan Lake. The New Hope Creek arm has an average
hydraulic retention time of 418 days” (2)
199 NPDES permitted dischargers to Jordan Lake watershed (2)

Historical Water Quality Problems
“Historical water quality problems observed at Jordan Lake include low dissolved oxygen, elevated nutrients, elevated chlorophyll a, algae blooms, taste and odor problems at the Cary/Apex
water intake, and fish kills” (3)
1995 and 1996 complaints of taste and odor of Cary’s water “were associated with the blue-green alga Anabaena spirodes and the diatoms Melosira spp.” (3)
Plot of historic NCTSI scores show that although the TSI scores have decreased over time, the lake remains eutrophic (16)

Nutrients
“Nutrients were found in elevated amounts frequently during the Jordan Lake Study.” Ammonia was found in elevated amounts ((0.05 mg/L) in 18% of total samples, TKN was found in elevated
amounts ((0.6 mg/L) in 15% of samples, NOx was greater than or equal to 0.05mg/L in 48% of samples, total phosphorus greater than or equal to 0.05 mg/L in 61% of samples, and orthophosphorus
was found in elevated amounts ((0.05 mg/L) in 19% of samples. Tables 3 and 4 on page 10 provide the maximum, minimum, and average nutrient values for each sampling location for 1996
and 1997 (9-10)
Algal Growth Potential Test (AGPT) was conducted for Haw River and Jordan Lake and the results indicated all five stations were limited for nitrogen (20)

Chlorophyll a
17% of chlorophyll samples collected in 1996 and 1997 exceeded standard. (13)
“The highest chlorophyll a values found each month were generally at the upstream stations in the Haw River and, especially, in the New Hope Creek arms of the lake” (13)
Figure 11 provides the frequency of chlorophyll a violations in last 5 years and period of record for six stations on Jordan Lake. The station with the highest percentage of exceedences
for the period of record (55%) is located on the Upper New Hope Arm. The station with the high percentage of violations in the last five years (57%) is located on the Morgan Creek
arm (17)
11% of pH samples above standard; 7/11 of these exceedences occurred on August 22, 1996 and were associated with algal blooms and elevated chlorophyll a levels (18)

Phytoplankton
“From 1984 to 1994, 17 algae blooms were documented in Jordan Lake and its tributaries by DWQ.” (21)
“Eighty-nine percent of samples collected [1996-1997] were determined to be algal blooms based on phytoplankton data.” Physical/chemical data also indicated bloom conditions (elevated
chlorophyll a, supersaturated epilimnion, and high pH) (21)

Metals
19% of copper values were greater than or equal to action level of 7 g/l (27)
17% of zinc values were greater than or equal to action level of 50 g/l (27)

Hoyer, M.V. and Jones, J. R. “Factors Affecting the Relation Between Phosphorus and Chlorophyll a in Midwestern Reservoirs.” Canadian Journal of Fisheries and Aquatic Sciences. 40
(2): 192-199. 1983.

Summary
This article discusses the relationship between phosphorus and chlorophyll a in midwest reservoirs. Variables for nitrogen, zooplankton abundance, hydrologic flushing, and inorganic
suspended solids were added to log-normal regressions of chlorophyll a on total phosphorus to reduce the error in this relation. The only variable that reduced the error term was inorganic
suspended solids. A multivariate equation that used the ratio of inorganic suspended solids to total phosphorus was developed to account for the effect of inorganic solids on the phosphorus-chlorop
hyll a relationship. This equation can be used to predict chlorophyll a concentrations in lakes with inorganic suspended solids.

Chlorophyll a – Phosphorus Relationships in Reservoirs
Investigators “have described a strong relationship between total phosphorus concentrations (spring or summer) and summer algal biomass measured by chlorophyll a” in natural lakes. “The
relation is weaker in reservoirs, but many reservoirs have algal levels in agreement with the phosphorus – chlorophyll a relation.”

“The slope and intercept values from the log-normal regressions of chlorophyll a on total phosphorus in the literature are surprisingly consistent, despite differences among the data
sets. However, deterministic predictions of chlorophyll a based on a change in lake phosphorus concentration are not always useful to lake managers because of variations in the relation.
… For example, a lake with 20 mg/m3 total phosphorus could have a chlorophyll a concentration ranging from 2 to 15 mg/m3 … Predictions are even less reliable in reservoirs with nonalgal
turbidities” (193)

Materials and Methods
“Surface waters of 96 reservoirs in Missouri and Iowa were sampled between May and September 1978-81.” The following parameters were sampled: total phosphorus, chlorophyll a, Secchi
transparency, total nitrogen, inorganic suspended solids, hydrologic flushing rate, and zooplankton (cladocerans, copepods, rotifers) (193-194)
“Total suspended solids were determined by filtering water through precombusted (550(C), preweighed, Whatman 934 AH glass fiber filters. Filters were weighed after drying at 103(C for
1 h. Inorganic suspended solids were determined after combustion at 550(C for 1 h. Organic suspended solids were determined by differences. Appropriate corrections were made for
blanks” (193)

Results and Discussions
There were no significant differences in the total phosphorus – chlorophyll a relation between the regressions equations for natural lakes and reservoirs. The authors concluded “mean
chlorophyll yield per until of total phosphorus and the variance about this yield are similar in many lakes and reservoirs” and assumed “the factors and mechanisms controlling variation
are similar in both lake types” (194)
“For our data there was no strong relation between nitrogen and chlorophyll a (R2=0.18). Thus, there was no reduction in the error sum of squares when nitrogen was regressed with total
phosphorus against chlorophyll a” (194)
“… in reservoirs with N:P ratios less than 10, nitrogen accounted for the same amount of variance as did phosphorus (R2=0.73 and R2=0.71, respectively)” (195)
Like, nitrogen, zooplankton and flushing rate did not reduce the error term when regressed with phosphorus against chlorophyll a (195-196)

Inorganic Suspended Solids
“Hergenrader and Hammer (1973) and Jones and Novak (1981) found that reservoirs with high concentrations of organic suspended solids may have as little as 10% of the chlorophyll that
would be predicted for a given total phosphorus concentration” (196)
The authors expected “to find a negative correlation between inorganic suspended solids and chlorophyll in our reservoirs. The simple coefficient, however, was positive and significant”
(196)
The authors hypothesized that the positive relation resulted from the high degree of intercorrelation among inorganic suspended solids, total phosphorus, and chlorophyll concentrations
(196).
The individual effect of total phosphorus and inorganic suspended solids on chlorophyll a were investigated with path analysis and partial correlation procedures. “The standard partial
regression path coefficient of chlorophyll a on inorganic solids when total phosphorus was held constant was negative (-0.31) and significant” (196)
“We found a significant negative correlation between residual chlorophyll a and the ratio of inorganic suspended solids to total phosphorus (I:TP) (r =-0.36) (196)
Based on the above findings, this equation was developed:
log chlorophyll a = -0.47 + 1.13 log tot phos – 1.03 (inorganic suspended solids/tot phos)
The above multivariate model “accounts for 7% more variance in our data than the univariate equation, and the 95% predictive confidence interval is reduced by 10%” (197)
Two hypothesis for how inorganic suspended solids reduce summer mean chlorophyll yield per unit phosphorus are “inorganic solids could decrease the biologically available phosphorus
in a lake” or they could “decrease light available for photosynthesis” (197)

B-24 Appendix B

Appendix B B-3

Appendix C.docInvestigating relationships between algal communities and water quality in Jordan lake is complex because of the amount of data available. These include algal species
counts and biovolumes that occur together with water quality data for the same sampling station and date (within one day) in about 460 samples. Some simplifying was necessary. Only
biovolumes summed for each algal division (phyla, for the most part) were analyzed. Only field measurements and nutrient data were included. The approach was to derive principal component
scores for algae and water quality separately for each date or station.

The principal component procedure found a series of gradients (principal components) in the algal community and, separately, in the water quality data. This was first done by date,
so that each date had a series of principal component scores. In general, more than 70 percent of the variation in the data was explained by the first 4 water quality components (gradients)
and the first 3 algal components. Analysis was restricted to these components. Combining the sets of scores by date, correlations could be calculated between the algal and water quality
scores. Correlating several scores by nearly a hundred dates resulted in about 900 comparisons. The frequencies with which correlations were significant were related to the significance
level chosen in such a way that chance appeared to be the most important factor. For example, at the 95 percent significance level 43 significant correlations were seen, while something
like 45 (0.05 X 900) may be expected by chance.

This approach was also taken for each station. This resulted in fewer correlation comparisons (about 80) being calculated between algal and water quality principal components scores.
At the 95 percent significance level 12 significant correlations were seen, while about 4 (0.05 X 80) may be expected by chance. Because of these results, the analysis by date was
set aside and the analysis by station was carried forward.

Where there was a significant correlation between station scores for a water quality component and an algal component, the scores were plotted (Attachment A). The first graph in Attachment
A has the second algal component (PPRIN2) for station B2451500 plotted on the first water quality component (WPRIN1) for this station. Each water quality component had separate scores
for each variable (temperature, phosphate, etc.), indicating which variable had the highest correlation with the component. For the first graph, total inorganic nitrogen (TIN) and
nitrate + nitrite (NOx) had the highest positive correlations with this component, and temperature (temp), pH, and chlorophyll a (chla) had the most negative correlations. This means
that points on the right side of the graph have higher NOx concentrations and lower temperatures than those on the left. Gradients in the algal data were treated similarly. The notation
on the vertical scale of the first graph shows that points, which represent dates, in the upper part of the graph are associated with low blue-green algal biovolume (cya) and high biovolumes
of prymnesioids (pry) and euglenoids (eug). The inferred negative relationship between blue-greens and NOx and the positive relationship between blue-greens and temperature are the
most frequently seen in the set of 12 graphs.

The symbols plotted in the graphs indicate each date, as seen in Attachment B. The simplest way to interpret them is to realize that capital letters represent spring and summer (April
through September) and lower case letters represent fall and winter. The complexity of the scheme used to identify each year and season was required by limitation to a single character.

Because the relationships between blue-greens and temperature and blue-greens and NOx were seen at most stations, they were investigated further. In fact, regressions predicting blue-green
biovolume were calculated using all the nutrient and field measurements. In a stepwise multiple regression procedure, temperature was the best predictor, explaining about 50 percent
of the variability in the data (ln(b-g biovolume [mm3/M3]) = 0.345 X temperature [(C] - 1.85; p>F = 0.0001; N = 353). Other water quality measurements had minimal contributions in
this procedure. Because NOx and temperature were correlated, a separate, simple regression was calculated for NOx predicting blue-green biovolume (ln(b-g biovolume [mm3/M3]) = 2.00
- 1.30 X ln(NOx concentration [(g/L]); p>F = 0.0001; N = 353). About 35 percent of the variation in the data was explained by the NOx relationship. (Scatter plots for these two relationships
are in Attachment C.) Although not as good a fit as the temperature prediction, the negative relationship with NOx perhaps has more direct biological meaning. As inorganic nitrogen
becomes less available in the photic zone during summer, blue-greens may be favored because many of them can fix atmospheric nitrogen.

ATTACHMENT A
Graphs for Significant Correlations Between Algal and Water Quality Principal Component Scores

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1356
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B2451500 ----------------------------------------

Plot of PPRIN2*WPRIN1. Symbol is value of PLOTSYM.

3 ˆ
‚
‚
‚
‚ a
‚ d b
2 ˆ C a
‚ D d
- cya ‚ C D
‚ q A
+ pry, eug ‚ D C A c
‚ o E
1 ˆ O d a b
‚ o ppb o
‚ s c
PPRIN2 ‚ T B P B
‚ A
‚ Q
0 ˆ C
‚
‚ E r q O
‚ Q Q q
‚ U p e
‚ T G
-1 ˆ r e
‚
‚ R B O e
‚ PE R F u G S
‚ R U F r
‚ P
-2 ˆ
‚ S t
‚
‚ Y XX Y L J Y K X K
‚
‚
-3 ˆ
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-6 -4 -2 0 2 4

WPRIN1 + TIN, NOx - temp, pH, chla

NOTE: 1 obs had missing values. 1 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1359
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B2451500 ----------------------------------------

Plot of PPRIN2*WPRIN3. Symbol is value of PLOTSYM.

3 ˆ
‚
‚
‚
‚ a
‚ b d
2 ˆ a C
‚ d D
- cya ‚ C D
‚ q A
+ pry, eug ‚ CcA
‚ o E
1 ˆ b O a d
‚ oo p p
‚ s c
PPRIN2 ‚ PT B B
‚ A
‚ Q
0 ˆ C
‚
‚ r qO E
‚ qQ Q
‚ U e p
‚ T
-1 ˆ r e
‚
‚ Oe RB
‚ uS G F P E
‚ UF R r
‚ P
-2 ˆ
‚ S t
‚
‚ YX XJK L Y ZY
‚
‚
-3 ˆ
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-4 -2 0 2 4 6

WPRIN3 + PO4, TP - TKN, TON

NOTE: 1 obs had missing values. 6 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1362
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B2453000 ----------------------------------------

Plot of PPRIN2*WPRIN1. Symbol is value of PLOTSYM.

2 ˆ U
‚
‚ R R u a
‚ Y CT Y
‚ O J
‚ Z Y SO
1 ˆ L P o t o
‚ SF K U r
+ cya, ‚ Z E D
chl ‚ Q G
‚ D
- pry ‚ p D O F s
0 ˆ X R p o
‚ A qq
‚ X Q X CQ A r
PPRIN2 ‚ B C C B
‚ E a a e
‚ T A p c
-1 ˆ PG
‚ E d b d
‚ q e
‚ c rd
‚ B
‚
-2 ˆ
‚
‚
‚
‚ e
‚ b
-3 ˆ b
‚
‚
‚
‚
‚ K
-4 ˆ
Šƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-6 -4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp, pH

NOTE: 3 obs had missing values.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1363
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B2453000 ----------------------------------------

Plot of PPRIN3*WPRIN1. Symbol is value of PLOTSYM.

PPRIN3 ‚
‚
3 ˆ
+ cry ‚
‚ K
- eug ‚
2 ˆ
‚
‚
‚ R B u
1 ˆ Z Y Y R J T D
‚ Y Q C C t G B
‚ X C Q qr E d a dr
‚ X Z p D U q F B p c d
0 ˆ L SF O E s q
‚ K S e e e
‚ Q U PG D p b
‚ O A r
-1 ˆ A o
‚ a
‚ O b ob
‚
-2 ˆ
‚ a
‚
‚
-3 ˆ
‚
‚
‚ A
-4 ˆ
‚ T o
‚
‚
-5 ˆ
‚
Šƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-6 -4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp, pH

NOTE: 3 obs had missing values. 6 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1368
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B3959000 ----------------------------------------

Plot of PPRIN2*WPRIN1. Symbol is value of PLOTSYM.

PPRIN2 ‚
4 ˆ a
‚
‚
‚
‚ a
3 ˆ a
- cya ‚
‚
+ eug ‚
‚
2 ˆ
‚
‚ b
‚
‚ E e
1 ˆ D cd
‚ P B ed r
‚ K E D G U CP pSC o b e
‚ E Q C q D q p Bp q
‚ X RA b o d
0 ˆ r
‚ A F o
‚ X A R s S r
‚ R T t
‚ O Z
-1 ˆ O
‚ X B O uQ
‚ T Q
‚ G
‚ L
-2 ˆ Y U K
‚
‚
‚ J
‚
-3 ˆ Y
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp

NOTE: 1 obs had missing values. 4 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1370
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B3959000 ----------------------------------------

Plot of PPRIN1*WPRIN2. Symbol is value of PLOTSYM.

PPRIN1 ‚
2 ˆ R
‚
+ cry, chl ‚ p r
‚ p D P T
‚ K p E CZ C Q
1 ˆ P Y t
‚ dU S D Q
‚ d B E Xc F P o R
‚ E q q s r r o
‚ L e X AC Z S
0 ˆ KY Q
‚ Au O
‚ b B d O D T
‚ O Z
‚ B o
-1 ˆ
‚ b
‚
‚ U
‚
-2 ˆ
‚
‚ A
‚
‚
-3 ˆ J a
‚
‚ b
‚ X
‚
-4 ˆ
‚ Y
‚
‚ a
‚
-5 ˆ a
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-4 -2 0 2 4 6

WPRIN2 +TKN, TON, chla

NOTE: 1 obs had missing values. 6 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1373
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B3968000 ----------------------------------------

Plot of PPRIN1*WPRIN1. Symbol is value of PLOTSYM.

PPRIN1 ‚
2 ˆ
‚
+ cya, ‚ O r r
cry ‚ Y X KY LXROYX S J K q p
‚ R O GRQ sq F
‚ D U C p P p
‚ EPP B S T t E o dr
0 ˆ D C
‚ T FA Q B B o e
‚ c b
‚ b e do A
‚ E d
‚
‚ A
-2 ˆ
‚ b
‚
‚
‚
‚
‚
-4 ˆ
‚ a
‚ a
‚
‚
‚
‚
-6 ˆ
‚
‚ a
‚
‚
‚
‚
-8 ˆ
Šƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-6 -4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp

NOTE: 10 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1374
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B3968000 ----------------------------------------

Plot of PPRIN2*WPRIN1. Symbol is value of PLOTSYM.

PPRIN2 ‚
2 ˆ
‚
+ chr, ‚ C
eug ‚ D a
‚ q q d
1 ˆ B E G F p a
‚ P E C p B e
‚ G S p re
‚ D OQ r b Aq A
‚ E C T B e
0 ˆ
‚ R s d d
‚ P Q b
‚ O o b
‚ R RA P
-1 ˆ F S a
‚ o
‚ O t
‚ T
‚ Y X KY LX YX u J K
-2 ˆ
‚
‚
‚
‚
-3 ˆ o
‚
‚
‚ U
‚
-4 ˆ
‚
‚
‚
‚
-5 ˆ Q
Šƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-6 -4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp

NOTE: 6 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1378
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B3968000 ----------------------------------------

Plot of PPRIN1*WPRIN4. Symbol is value of PLOTSYM.

PPRIN1 ‚
2 ˆ
‚
+ cya, ‚ O Q r r
cry ‚ Z X Ku R O Y X Xp SK qJ Z
‚ UO Q G G R q R F
‚ P D p U C q p
‚ E tP TP C SE dB o r
0 ˆ DC
‚ e A B e TF o Q
‚ b c
‚ e b d o A
‚ d E
‚
‚ A
-2 ˆ
‚ b
‚
‚
‚
‚
‚
-4 ˆ
‚ a
‚ a
‚
‚
‚
‚
-6 ˆ
‚
‚ a
‚
‚
‚
‚
-8 ˆ
Šƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-3 -2 -1 0 1 2 3

WPRIN4 + NH3 - TP

NOTE: 6 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1380
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B3980000 ----------------------------------------

Plot of PPRIN2*WPRIN1. Symbol is value of PLOTSYM.

PPRIN2 ‚
‚
1.5 ˆ
‚
‚
‚ A a
1.0 ˆ
+ chl, chr ‚
‚
- eug ‚
0.5 ˆ a
‚
‚
‚
0.0 ˆ
‚
‚
‚
-0.5 ˆ a
‚
‚
‚
-1.0 ˆ
‚
‚
‚
-1.5 ˆ
‚
‚
‚
-2.0 ˆ
‚
‚ Z P LOY APKOYO ZPX J p A B opb c Bo o b b
‚
-2.5 ˆ
‚
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-4 -2 0 2 4 6

WPRIN1 + TN, NOx - temp

NOTE: 6 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1385
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B4010000 ----------------------------------------

Plot of PPRIN1*WPRIN1. Symbol is value of PLOTSYM.

PPRIN1 ‚
4 ˆ U
‚
+ cry, cya ‚
‚
- bac ‚
‚
‚
2 ˆ XZ Y KXL KX J u
‚ BQ s
‚ SO r
‚ O t
‚ O Q R Q B
‚ C C D D q o p r
‚ D E q F q
0 ˆ U p B o p o
‚ F S P A d d b
‚ P C e r
‚ A A b c e
‚ E d
‚ c b
‚
-2 ˆ T
‚ E
‚
‚
‚
‚
‚
-4 ˆ a
‚
‚ a
‚ a
‚
‚
‚
-6 ˆ
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp

NOTE: 2 obs had missing values. 9 obs hidden.

CORRELATIONS BETWEEN date PC SCORES FOR field measurements/nuts AND algal divs 1386
09:17 Friday, September 1, 2000
SCATTER PLOT OF DATES BETWEEN PC SCORES FOR NUTS-FLD MEAS AND ALGAL DIVISIONS

--------------------------------------- STORET=B4010000 ----------------------------------------

Plot of PPRIN2*WPRIN1. Symbol is value of PLOTSYM.

PPRIN2 ‚
‚
2 ˆ
‚ E
+ chr, bac, ‚
pry ‚ Q E C D E D ed d b p e
‚ ODC q pqB q e
‚ A c b
‚ G T FR A
0 ˆ QR tC s G c F d r
‚ O pP o
‚ P r
‚ B A r o
‚ U o b
‚ XZ YS KXL KX J u
‚ S a
-2 ˆ a a
‚ Q
‚
‚ T
‚
‚
‚
-4 ˆ
‚
‚
‚
‚
‚
‚
-6 ˆ U
‚
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-4 -2 0 2 4 6

WPRIN1 + TIN, NOx - temp

NOTE: 2 obs had missing values. 6 obs hidden.

ATTACHMENT B
Plot Symbols for PC Graphs

Year Month 1 2 3 4 5 6 7 8 9 10 11 12 83 a A O o 84 b B P p 85 c C Q q 86 d D R r 87 e E S s 88 f F T t 89 g G U u 96 j
J X x 97 k K Y y 99 l L Z z

ATTACHMENT C
Scatter Plots for Blue-Green Biovolume and Water Quality Measurements

Plot of CYA*LNNOX. Legend: A = 1 obs, B = 2 obs, etc.

CYA ‚
‚
15 ˆ
‚
‚
‚
‚
‚
‚ A A
10 ˆ B
‚ P A A A
‚ Z A A A AA A A A
‚ R A A B B C A A A B AB A BB A
‚ U A C A A AA AA BB AA AB
‚ O B B A A B AA AB CACAAAA A
‚ E A A B A AB A AA B AAAABAB CB A
5 ˆ H B ABBB AA B A
‚ B A A A A A AE CB CAA AAAA A
‚ A B ABBAA DAA A A
‚ A AA CBB AB A
‚ A A AA A ACA A ABCB A
‚ AA BA BAA AA A
‚ A B AA AA
0 ˆ A A A A
‚ A A AA
‚ A
‚
‚
‚
‚
-5 ˆ
‚
‚
‚
‚
‚
‚ A A A CDAAAA A
-10 ˆ
‚
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
-5 -4 -3 -2 -1 0 1 2

LNNOX

NOTE: 421 obs had missing values.

Plot of CYA*TEMP. Legend: A = 1 obs, B = 2 obs, etc.

CYA ‚
‚
15 ˆ
‚
‚
‚
‚
‚
‚ A A
10 ˆ A A
‚ A AAA CBCFA
‚ A BDAA ADBECCABBB
‚ AA A B BBAAAB D BAAD DCBCB A
‚ AB ABA A A A A CB A AAACEBBBAB
‚ A AAB BB CAB BEAA C AABBAAB B A
‚ A A A BA BBB A ABC ACAA A A A A AAAA
5 ˆ A A AA B AAA AA AB AA A B A A A
‚ A AA A AA A AAAB CB A B A AA AA B A
‚ B A B BA A AA B AA A AA
‚ AA AAA BABA A B
‚ A A A AA A AAAABB A AB A A
‚ A A A A A ABAA A A
‚ B A A B A
0 ˆ A A A A
‚ AA A A
‚ A
‚
‚
‚
‚
-5 ˆ
‚
‚
‚
‚
‚
‚ B AA BAB AA AA B
-10 ˆ
‚
Šƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒƒƒƒˆƒƒ
0 5 10 15 20 25 30 35

TEMP

NOTE: 424 obs had missing values.

C-18 Appendix C

Appendix C C-19

Appendix D/Bynum FLUX.doc

FLUX Analysis of Nutrient Loading
on the Haw River near Bynum
August 1983 – December 1999

April 26, 2001

This document has been prepared by Tetra Tech, Inc. for the
Jordan Lake Nutrient Response Modeling Project.
Funding for this project is provided by:

City of Burlington
City of Graham
City of Greensboro
Town of Mebane
Orange Water and Sewer Authority
Town of Pittsboro
City of Reidsville

Table of Contents

Table of Contents 2
Introduction / Description of Data Sources 3
Total Nitrogen 4
Total Kjeldahl Nitrogen 27
Ammonia 41
Total Phosphorus 53

Introduction

To properly calibrate and verify the Jordan Lake Nutrient Response Model, complete or continuous estimates of tributary nutrient loads must be available or developed. Observed nutrient
data from tributaries consists of point-in-time measurements of concentration, along with continuous estimates of flow from USGS gaging stations. Converting such point-in-time concentrations
to continuous load estimates presents a number of technical challenges because concentration and flow are typically correlated instead of being independent of one another. A variety
of methods are available to address this problem, but no one method is superior for all situations. The US Army Corps of Engineers Waterways Experiment Station produced the FLUX program
to evaluate and compare multiple methods of estimating nutrient loads from tributary concentration data. Tetra Tech used FLUX to analyze nutrient loads for each of the four monitored
tributaries to Jordan Lake. These tributaries are (a) the Haw River; (b) New Hope Creek; (c) Morgan Creek; and (d) Northeast Creek. This report presents the results of the FLUX analysis
for the Haw River near Bynum.

The FLUX analysis required input files containing mean daily flows and sample data consisting of measured concentrations of various nutrient species. The daily mean flows of the Haw
River were obtained from the USGS (gage 02096960) and the water quality monitoring data were a combination of USGS data collected at the flow gage and North Carolina Division of Water
Quality data. Water quality data for both agencies is collected at the US Highway 15/501 bridge crossing near Bynum, NC. The flow gage is located on the right bank, 1.1 mi. downstream
of US 15/501.

Data selected for this analysis cover the period of August 1, 1983 through December 31, 1999. The end date was selected to correspond to the last samples available from calendar year
1999 (the last year for which complete flow and water quality data were available). The start date for the analysis was selected to provide approximately 6,000 mean daily flows (the
maximum number of flows that can be analyzed by the FLUX code).

Table 1 below compares the entire flow record from August 1, 1983 – December 31, 1999 to the flow record for the dates on which nutrient measurements were made (sample flows).

Table 1. Flow Data Comparison
Number of Measurements Flow Range
(cfs) Mean Value
(cfs) Median Value
(cfs) Entire Flow Record 5,997 .18 – 58,000 1,263 512 Flow on days when nutrient measurements were taken 231 1.5 – 25,000 1,992 620
Table 1 indicates that the sample flow values have a similar median and average to the entire flow record. However, the sample values do not include values from the high end of the
flow range.

Total Nitrogen
FLUX was used to analyze 231 total nitrogen (TN) samples from August 1, 1983 through December 31, 1999. The TN load for this period was calculated using six estimation methods (Table
1A). Method 6 was used in the subsequent analysis because it has the lowest coefficient of variation (CV), .043, of all the methods.

The TN load was also calculated for the data from January 1, 1990 to December 31, 1999 (Table 1B). The lowest CV for the load calculation was .050, using Method 6 (REG-3).

The mean and standard deviation of the six methods calculating the loads for the entire period (8/1/83 – 12/31/99) are shown in Figure 1.

Concentration versus Date
Figure 2 shows TN concentration versus date for the entire period. This figure illustrates the average concentration is relatively constant over the 16-year period. Table 2 provides
a statistical summary of the relationship between TN concentration and date. Figure 3 illustrates the residuals versus date for Method 6. There is a downward slope in the regression
residuals, suggesting that the relationship between TN and flow may have changed over time.

Concentration versus Flow
Figure 4 illustrates TN concentration versus flow using method 6 and Table 3 provides the statistics associated with this graph. Figure 4 demonstrates that the concentration decreases
with flow. Figure 5 illustrates the residuals versus flow.

The slope of the relationship between concentration and flow is statistically significant; however, the residuals do not have as significant slope versus flow. This suggests that the
estimation method adequately accounts for the flow-concentration relationship.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in TN concentration over time. The first stratum included data from August 1, 1983 – December
31, 1989 while the second stratum included data from January 1, 1990 – December 30, 1999. The load calculations for the six estimation methods are show in Table 4. Method 6 again
had the lowest CV, .039. Since the CV for this method is slightly less than CV for this method without stratification, the concentration was graphed against flow (Figure 6).

The data were then stratified by date using the following ranges: August 1, 1983 – December 31, 1988, January 1, 1989 – December 31, 1994 and January 1, 1995 – December 30, 1999. The
load calculations for this stratification scheme are shown in Table 5. The CV for Method 6 was the lowest, .036. Figure 7 illustrates TN concentration versus flow the three.

Both date stratification schemes suggested a small decrease in flow-normalized TN concentration over time.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 6 illustrates the load calculations for each estimation method. The CV for Method 6, 0.041,
was the lowest of the estimation methods. Figure 8 illustrates concentration versus flow for the seasonal stratification. The graph illustrates a slight increase in TN concentrations
in the growing season (May – September), particularly with the higher flows. However, there have not been very many TN measurements during high flows in the growing season.

The seasonal stratification was also conducted on the abridged record, from January 1990 to December 1999. Table 7 shows the load calculation for the seasonal stratification on this
period. Method 6 had the lowest coefficient of variation, .048. Concentration vs. flow for this date stratification is shown in Figure 9.

Flow Stratification
The data were also stratified by flow, testing 900 cubic hectometers/year (1,011 cfs) and 1,200 cubic hectometers/year (1,348 cfs) as the cutoff value, based on visual observation of
the concentration vs. flow plot (Figure 4). For the stratification of flows greater than and less than 900, the lowest CV was .047 (Method 4) while the CV for Method 6 was .048 (not
shown). For a stratification using 1,200 as the cutoff, the lowest CV was .046 (Method 4) while the CV for Method 6 was .050 (not shown). Since the lowest coefficients in the two
flow stratification schemes are higher than the CV for Method 6 with no stratification, the data suggests no improvement in load estimation is achieved through the use of flow stratification.

Conclusions
Method 6 was the best for all stratification schemes except flow stratification
TN concentration appears to have decreased slightly over time for the period between August 1983 and December 1999. However, the USGS reported no statistically significant trends in
the TN concentration at the Haw River near Bynum from 1983 – 1995 using the non-parametric seasonal Kendall test.
TN concentration varies slightly with seasons (growing and non-growing). However, there were few measurements of TN concentrations in high flows during the growing season. Also, the
relationship was weaker when the shorter record was used.
Using an abridged record (1/90 – 12/99 instead of 8/83 – 12/99) led to higher CVs for all stratification schemes (and no stratification).
There was no improvement when the data were stratified by flow because the relationship between flow and concentration is adequately represented by the regression method.

Recommended Method and Stratification
Based on the conclusions and findings above, the recommended procedure for estimating TN loads for the Haw River near Bynum is:
Use Method 6 for the entire period (8/83 – 12/99) with no stratification

Load Calculations
Table 8 shows the annual TN loads for the entire period using the recommended procedure, Method 6 and no stratification. Note that FLUX provides both direct ‘model’ outputs and a modified
estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges
not accounted for by date stratification. Also note that results from 1983 reflect only a partial year, and are therefore not directly comparable to annual totals from subsequent full
years.

Table 1A: Calculation of Total Nitrogen (TN) Load on the Haw River near Bynum using Six Estimation Methods (8/83 – 12/99)

Table 1B: Calculation of TN Load on the Haw River near Bynum using Six Estimation Methods (1/90 – 12/99)

Figure 1: Mean and one Standard Error for Six Estimation Methods

Figure 2: TN Concentration (µg/l) versus Date

Table 2: Statistics for Graph of TN Concentration versus Date

Figure 3: Plot of TN Residuals versus Date
Figure 4: TN Concentration (µg/l) versus Flow (hm3/yr)

Table 3: Statistics of TN Concentration versus Flow

Figure 5: TN Residuals versus Flow (hm3/yr)

Table 4: Calculation of TN Loads with Six Estimation Methods using Date Stratification (83-89 and 90-99)

Figure 6: TN Concentration (µg/l) versus Flow (hm3/yr) for Date Stratification
Strat 1= 83-89; Strat 2=90-99
Table 5: TN Load Calculations with Date Stratification (83-88, 89-94, 95-99)
Figure 7: TN Concentration (µg/l) versus Flow (hm3/yr) for Date Stratification (83-88, 89-94, 95-99) Strat 1= 83-88; Strat 2=89-94; Strat 3=95-99

Table 6: Load Calculations for TN using Season Stratification
Figure 8: TN Concentration (µg/l) versus Flow (hm3/yr) for Seasonal Stratification
Strat 1= Oct-April; Strat 2= May-September
Table 7: TN Load Calculations using Data from 1/90 – 12/99 and Seasonal Stratification
Figure 9: TN Concentration (µg/l) vs Flow (hm3/yr) for Seasonal Stratification, 1990-1999
Table 8: Annual TN Loads August 1983 – December 1999
Nitrate/Nitrite
Using FLUX, the nitrate/nitrite (NOx) load from August 1983 – December 1999 was calculated using six estimation methods (Table 1A). Method 6 was used in the subsequent analysis because
it has the lowest coefficient of variation (CV), .039, of all the methods.

The NOx load was also calculated for the data from January 1, 1990 to December 31, 1999 (Table 1B). The lowest CV for this load calculation was .059, using Method 6 (REG-3).

The mean and standard deviation of the six methods calculating loads for the entire period (8/1/83 – 12/31/99) are shown in Figure 1.

Concentration versus Date
Figure 2 shows NOx concentration versus date for the entire period. This figure illustrates the average concentration is relatively constant over the 16-year period. Table 2 provides
a statistical summary of the relationship between NOx concentration and date. Figure 3 illustrates the residuals versus date. The regression residuals are basically constant throughout
the 16-year period, but slightly decrease with time.

Concentration versus Flow
Figure 4 illustrates NOx concentration versus flow using Method 6 and Table 3 provides the statistics associated with this graph. As with TN, the concentration decreases with flow.
Figure 5 illustrates the residuals versus flow. Results suggest that the regression approach adequately describes the flow-concentration relationship.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in NOx concentrations over time. The first stratum included data from August 1, 1983 – December
31, 1989 while the second stratum included data from January 1, 1990 – December 30, 1999. The load calculations for the six estimation methods are show in Table 4. Method 6 again
had the lowest coefficient of variation (CV), .042. Since this CV is higher than the CV for no stratification and Figure 2 illustrates no trend in NOx concentration with time, there
is no need to stratify by data and no further analysis was conducted on date stratification.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 6, 0.035,
was the lowest of the estimation methods. Figure 6 illustrates concentration versus flow for the seasonal stratification. The graph illustrates a slight increase in NOx concentrations
in the growing season (May-September), particularly with the higher flows. However, there have not been very many NOx measurements during high flows in the growing season.

Flow Stratification
The data were also stratified by flow, testing 900 cubic hectometers/year (1,011 cfs) and 1,200 cubic hectometers/year (1,348 cfs) as the cutoff value, based on visual observation of
the concentration vs. flow plot (Figure 4). For the stratification of flows greater than and less than 900, the lowest CV was .039 (Methods 4 and 6) (Table 6). For a stratification
using 1,200 as the cutoff, the lowest CV was .039 (Method 4) while the CV for Method 6 was .041. Figure 7 illustrates concentration versus flow for the flow stratification with 900
as the cutoff. Figure 7 illustrates the two strata have almost the same slope, but the downward regression line may be slightly steeper for the higher flows and there is a small jump
between the two trendlines. However, there is no need to stratify by flow because the general trend of decreasing concentration with increasing flow is the same for both strata and
the CV for the flow stratification schemes were the same as the CV for no stratification.

Conclusions
Method 6 was the best for all stratification schemes (and no stratification) except the flow stratification at 1,200 cubic hectometers/year
There is no clear change in NOx concentration over time. This is consistent with the USGS’s finding that there were no statistically significant trends in the NOx concentration at the
Haw River near Bynum from 1983 – 1995.
NOx varied with seasons (growing and non-growing). Although the trendlines crossed each other, the seasonal stratification lowered the CV.
Using an abridged record (1/90 – 12/99 instead of 8/83 – 12/99) led to a higher CV for no stratification. Since there was no significant correlation between NOx concentration and time,
there was no need to use the abridged record for the stratification schemes. Therefore, the maximum amount of concentration measurements (within the flow data constraints of FLUX)
could be used to analyze potential stratification schemes.
There was no improvement when the data were stratified by flow because the relationship between flow and concentration (decreasing concentrations with increasing flows) is adequately
explained by the regression model.

Recommended Method and Stratification
Based on the conclusions and findings above, the recommended procedure for estimating NOx loads for the Haw River near Bynum is:
Use Method 6 for the entire period (8/83 – 12/99) with seasonal stratification

Load Calculations
Table 7 shows the annual loads for NOx for the entire period with Method 6 and no stratification. Table 8 shows the annual NOx loads obtained by using the recommended procedure: Method
6, seasonal stratification, and the entire period (8/83 – 12/99). Note that FLUX provides both direct ‘model’ outputs and a modified estimator that includes interpolation on residuals.
The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges not accounted for by date stratification.
Also note that results from 1983 reflect only a partial year, and are therefore not directly comparable to annual totals from subsequent full years.

Table 1A: Calculation of NOx Load on the Haw River near Bynum using Six Estimation Methods (83-99)
Table 1B: Calculation of NOx Load on the Haw River near Bynum using Six Estimation Methods (90-99)

Figure 1: Mean and one Standard Error for Six Estimation Methods
Figure 2: NOx Concentration (µg/l) versus Time
Table 2: Statistics for NOx Concentration versus Time

Figure 3: NOx Residuals versus Date
Figure 4: NOx Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for NOx Concentration versus Flow
Figure 5: NOx Residuals versus Flow (hm3/yr)

Table 4: NOx Load Calculation with Date Stratification

Table 5: NOx Load Calculation with Season Stratification

Figure 6: NOx Concentration (µg/l) versus Flow (hm3/yr) for Season Stratification
(Strat 1 = Oct-April; Strat 2=May-Sept)
Table 6: NOx Load Calculations for Flow Stratification at 900 hm3/yr

Figure 7: NOx Concentration (µg/l) versus Flow (hm3/yr) for Stratification at 900 hm3/yr
Table 7: Annual NOx Loads, No Stratification
Table 8: Annual NOx Loads, Seasonal Stratification

Total Kjeldahl Nitrogen
FLUX was used to analyze 231 total Kjeldahl nitrogen (TKN) samples from August 1, 1983 through December 31, 1999. The TKN load for this period was calculated using six estimation methods
(Table 1A). Method 6 was used in the subsequent analysis because it has the lowest coefficient of variation (CV), .065, of all the methods.

The TKN load was also calculated for the data from January 1, 1990 to December 31, 1999 (Table 1B). The lowest CV for the load calculation was .066, using Method 6 (REG-3).

The TKN load was also calculated for the data from January 1, 1986 to December 31, 1999 (Table 1C) because Figure 2 illustrates high concentrations of TKN from 1983-1985. The lowest
CV for this load calculation, .055, was for Method 6. An examination of the flow record during this period showed that the flows were only slightly above average during this period
and there were no flows over 2500 cfs during this time. (The average flow from 8/1/83-12/31/85 was 1440 cfs compared to the average flow of 1302 cfs from 6/26/73 – 6/25/00). Since
these unusually high concentrations occurred at the beginning of the record and removing them from the record decreased the CV by .01, the subsequent analysis will be conducted with
data from 1/1/86 – 12/31/99 unless otherwise noted.

The mean and standard deviation of the six methods calculating the loads for the period from 1/1/86 – 12/31/99 are shown in Figure 1.

Concentration versus Date
Figure 2 shows TKN concentration versus date for the entire period (8/1/83-12/31/99). This figure illustrates that the average concentration is relatively constant over the 16-year period
but there are some relatively high concentrations between 1983 and 1985. Table 2 provides a statistical summary of the relationship between TKN concentration and date. Figure 3A illustrates
the residuals versus date for 8/1/83 – 12/31/99 and Figure 3B illustrates the residuals versus data for 1/1/86 – 12/31/99 for Method 6. Although there is a downward slope in the regression
residuals in both figures, the slope is much less in period from 1/86-12/99.

Concentration versus Flow
Figure 4 illustrates TKN concentration versus flow using Method 6 and data from 1/86 – 12/99, and Table 3 provides the statistics associated with this graph. Figure 5 illustrates the
residuals versus flow; the residuals are not correlated with flow.

Date Stratification
The data from 1986 on were stratified according to date to determine if there were any detectable trends in TKN concentration over time. The first stratum included data from January
1, 1986 – December 31, 1992 while the second stratum included data from January 1, 1993 – December 30, 1999. The load calculations for the six estimation methods are show in Table
4. Method 6 again had the lowest CV, .059. Since this CV for this stratification is greater than the CV for no stratification and Figure 2 illustrates no relationship between date
and concentration, there is no need for date stratification after 1985.

Seasonal Stratification
The data (1986-1999) was stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method
6, 0.054, was the lowest of the estimation methods. Figure 6 illustrates concentration versus flow for the seasonal stratification. Figure 6 illustrates the trendlines are very similar
for the two strata and that the concentration varies with flow more than it varies with season.

Flow Stratification
The data were also stratified by flow, testing 900 cubic hectometers/year (1,011 cfs) and 1,100 cubic hectometers/year (1,235 cfs) as the cutoff value, based on visual observation of
the concentration vs. flow plot (Figure 4). For the stratification of flows greater than and less than 900, the lowest CV was .051 (Method 4) while the CV for Method 6 was .056 (Table
6). For a stratification using 1,100 as the cutoff, the lowest CV was .050 (Method 4) while the CV for Method 6 was .055 (Table 7). Since the CV for a flow stratification with at
cutoff of 1100 using Method 4 was the lowest CV obtained in the TKN analysis, this scheme was used to plot concentration versus flow (Figure 7). Figure 7 clearly illustrates that TKN
concentrations first decrease with increasing flows (to approximately 1,100 cubic hectometers/year) and then increase with flow for flows greater than 1,100 cubic hectometers/year.

Flow and Seasonal Stratification
Since the data is dominated by flow stratification and may also have a seasonal stratification, the loads were calculated using both a combined seasonal and flow stratification scheme
(Table 8). The lowest CV for this stratification scheme (Strat 1 = less than 1,100, Oct-April; Strat 2= greater than 1,100, Oct-April; Strat 3= less than 1100, May-Sept; Stat 4= greater
than 1100, May-Sept) was .055 for Method 4. This is not a viable stratification scheme because the CV is higher than it is for flow alone and Strat 4 (growing season, high flows) only
contains 10 TKN measurements, compared with 40, 55, and 62 measurements in the other three strata. However, concentration versus flow using this stratification scheme is shown in Figure
8, which illustrates that the difference between seasons is more pronounced during high flows.

Conclusions
While Method 4 was best for load calculations with flow stratification, Method 6 was best for load calculations with seasonal stratification and for no stratification.
While there was no clear relationship date (a finding that is consistent with USGS findings), the CV was lowest when data from the years 1983-85 were not used in the analysis because
they had abnormally high TKN measurements. Using data from only the 1990’s not only increased the CV with respect to the record from 1986-99, but it also reduced the amount of data
available for analysis.
TKN concentration varied slightly with seasons (growing and non-growing). However, the use of a seasonal stratification either alone or in conjunction with a flow stratification did
not reduce the CV.
There is a significant improvement when the data were stratified by flow (at 1100 cubic hectometers/year) because the relationship between flow and concentration changed between low
and high flows

Recommended Method and Stratification
Based on the conclusions and findings above, the recommended procedure for estimating TKN loads for the Haw River near Bynum is:
Use Method 4 for the period 1/1/86 – 12/31/99 with flow stratification at 1100 cubic hectometers/year

Load Calculations
Table 9 shows the annual loads for TKN for 1/86-12/99 with Method 4 and no stratification. Table 10 shows the annual TKN loads obtained by using the recommended procedure: Method 4,
flow stratification at 1100, and data from 1/86 – 12/99. Note that FLUX provides both direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The
‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges not accounted for by date stratification.

Table 1A: Calculation of TKN Load on the Haw River near Bynum using Six Estimation Methods (8/83 – 12/99)
Table 1B: Calculation of TKN Load on the Haw River near Bynum using Six Estimation Methods (1/90 – 12/99)

Table 1C: Calculation of TKN Load on the Haw River near Bynum using Six Estimation Methods (1/86 – 12/99)
Figure 1: Mean and one Standard Error for Six Estimation Methods (86-99)
Figure 2: TKN Concentration (µg/l) versus Date (8/1/83 – 12/31/99)
Table 2: Statistics for TKN Concentration versus Date
Figure 3A: TKN Residuals versus Date (83-99)
Figure 3B: TKN Residuals versus Date (86-99)

Figure 4: TKN Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for TKN Concentration versus Flow

Figure 5: TKN Residuals versus Flow (hm3/yr)
Table 4: TKN Load Calculations for Date Stratification
(Strat 1 = 86-92; Strata 2 = 93-99)
Table 5: TKN Load Calculations for Seasonal Stratification

Figure 6: TKN Concentration (µg/l) versus Flow (hm3/yr) for Seasonal Stratification
(Strata 1=Oct-April; Strata 2=May-September)

Table 6: TKN Load Calculations for Flow Stratification at 900 hm3/yr

Table 7: TKN Load Calculations for Flow Stratification at 1,100 hm3/yr

Figure 7: TKN Concentration (µg/l) versus Flow (hm3/yr) with Flow Stratification at 1,100 hm3/yr

Table 8: TKN Load Calculations with Flow and Seasonal Stratification

Figure 8: TKN Concentration (µg/l) versus Flow (hm3/yr) with Flow and Seasonal Stratification
(Strat 1 = less than 1,100, Oct-April; Strat 2= greater than 1,100, Oct-April; Strat 3= less than 1,100, May-Sept; Stat 4= greater than 1,100, May-Sept)

Table 9: Annual TKN loads with Method 4 and no Stratification
Table 10: Annual TKN loads with Method 4 and Flow Stratification at 1,100 hm3/yr

Ammonia
FLUX was used to analyze ammonia (NH3) samples from August 1, 1983 through December 31, 1999. The NH3 load for this period was calculated using six estimation methods (Table 1A). The
lowest CV, 0.105, was obtained using Methods 2, 3, and 4. Method 4 was used in the subsequent analysis because it is the most widely used methods of the three mentioned above.

The NH3 load was also calculated for the data from January 1, 1990 to December 31, 1999 (Table 1B). The lowest CV for the load calculation was 0.121, using Method 2.

The NH3 load was calculated for the data from January 1, 1986 to December 31, 1999 (Table 1C) because Figure 2 shows the presence of elevated concentrations of NH3 from 1983-1985. The
lowest CV for this load calculation, 0.104, was for Method 4. An examination of the flow record during this period showed that the flows were only slightly above average during this
period and there were no flows over 2500 cfs during this time. (The average flow from 8/1/83-12/31/85 was 1440 cfs compared to the average flow of 1302 cfs from 6/26/73 – 6/25/00).
Since the high concentrations occurred at the beginning of the record, removing these values decreased the CV, and the TKN stratifications were performed on the data from 1/86-12/99,
the subsequent analysis on NH3 was conducted with data from 1/1/86 – 12/31/99 unless otherwise noted.

The mean and standard deviation of the six methods calculating the loads for the period from 1/1/86 – 12/31/99 are shown in Figure 1.

Concentration versus Date
Figure 2 shows NH3 concentration versus date for the entire period (8/1/83-12/31/99). This figure illustrates the average concentration is relatively constant over the 16-year period
but there are some relatively high concentrations between 1983 and 1985. Table 2 provides a statistical summary of the relationship between NH3 concentration and date. Figure 3 illustrates
the residuals versus date for 1/1/86 – 12/31/99. There is a slight downward slope to the regression residuals during this period.

Concentration versus Flow
Figure 4 illustrates NH3 concentration versus flow, using Method 4 and data from 1/86 – 12/99, and Table 3 provides the statistics associated with this graph. Figure 4 illustrates that
NH3 concentrations tend to increase with flow. Figure 5 illustrates the residuals versus flow; the residuals are not correlated with flow.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in NH3 concentration over time after 1985. The first stratum included data from January 1,
1986 – December 31, 1992 while the second stratum included data from January 1, 1993 – December 30, 1999. The load calculations for the six estimation methods are shown in Table 4.
Method 2 had the lowest CV, 0.118. Since the CV for this stratification is greater than the CV for no stratification and Figure 2 illustrates no relationship between date and concentration,
there is no need for date stratification.

Seasonal Stratification
The data (86-99) was stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method
5, 0.099, was the lowest of the estimation methods while the CV for Method 4 was the same as it was for no stratification, .104. Figure 6 illustrates concentration versus flow for
the season stratification using Method 5. Figure 6 illustrates the trendlines are very similar for the two strata and the difference between the seasons is less for high flows than
low flows. Since Method 5 is not widely used, the improvement from the seasonal stratification is minimum, and the differences between seasons is less for the critical high flows,
there is no need to stratify by season.

Flow Stratification
The data were also stratified by flow, testing 900 cubic hectometers/year (1,011 cfs) and 1,100 cubic hectometers/year (1,235 cfs) as the cutoff value, based on visual observation of
the concentration vs. flow plot (Figure 4). For the stratification of flows greater than and less than 900, the lowest CV was 0.097 for Method 4 (Table 6). For a stratification using
1,100 as the cutoff, the lowest CV was 0.100 for Method 4 (Table 7). Since the CV for a flow stratification with at cutoff of 900 using Method 4 was the lowest CV obtained in the NH3
analysis, this scheme was used to plot concentration versus flow (Figure 7). Figure 7 illustrates that NH3 increases with increasing flows for low flows and then increases less rapidly
for higher flows. It is possible that NH3 concentrations remain constant or even decrease slightly with high flows, but the lack of NH3 measurements at high flows make determining
a relationship difficult.

Conclusions
Method 4 was best for determining load calculations with no stratification and stratification by flow.
While there was no clear relationship with date (a finding that is consistent with USGS findings), the CV was lowest when data from the years 1983-85 were not used in the analysis because
they had abnormally high NH3 measurements. Using data from only the 1990’s not only increased the CV with respect to the record from 1986-99, but it also reduced the amount of data
available for analysis.
NH3 concentration varied slightly with seasons (growing and non-growing). However, the difference between seasons was less pronounced for high flows.
There is a significant improvement when the data were stratified by flow (at 900 cubic hectometers/year) but the general relationship between concentration and flow did not change between
the two strata.

Recommended Method and Stratification
Based on the conclusions and findings above, the recommended procedure for estimating NH3 loads for the Haw River near Bynum is:
Use Method 4 for the period 1/1/86 – 12/31/99 with stratification at 900 cubic hectometers/year

Although the general relationship between NH3 concentrations and flow did not change between the strata, the CV improved and stratification by flow was also performed for TKN. More
measurements of NH3 concentrations at high flows would be useful at solidifying the relationship between NH3 concentrations and high flows.

Load Calculations
Table 8 shows the annual NH3 loads obtained by using the recommended procedure: Method 4, flow stratification at 900 hm3/yr, and data from 1/86 – 12/99. Note that FLUX provides both
direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as
might be caused by changes in WWTP discharges not accounted for by date stratification.

Table 1A: Calculation of NH3 Loads using Six Estimation Methods (83-99)
Table 1B: Calculation of NH3 Loads using Six Estimation Methods (90-99)
Table 1C: Calculation of NH3 Loads using Six Estimation Methods (86-99)

Figure 1: Mean and one Standard Error for Six Methods (86-99)

Figure 2: NH3 Concentration (µg/l) versus Time (8/83-12/99)

Table 2: Statistics for NH3 Concentration versus Time
Figure 3: NH3 Residuals versus Date for 1986-1999
Figure 4: NH3 Concentration (µg/l) versus Flow (hm3/yr) for 1986-1999
Table 3: Statistics for NH3 Concentration versus Flow

Figure 5: NH3 Residuals versus Flow
Table 4: NH3 Load Calculations with Date Stratification (86-92 and 93-99)

Table 5: NH3 Load Calculations with Seasonal Stratification (Oct-Apr and May-Sept)

Figure 6: NH3 Concentration (µg/l) versus Flow (hm3/yr) for Seasonal Stratification for 1986-1999

Table 6: NH3 Load Calculations with Flow Stratification at 900 hm3/yr
Table 7: NH3 Load Calculations with Flow Stratification at 1,100 hm3/yr

Figure 7: NH3 Concentration (µg/l) versus Flow (hm3/yr) for Flow Stratification at 900 hm3/yr

Table 8: Annual NH3 Loads with Flow Stratification at 900 hm3/yr and Method 4

Total Phosphorus
FLUX was used to analyze total phosphorus (TP) samples from January 1, 1989 through December 31, 1999. Although data were available from earlier years, previous FLUX analysis confirmed
the 1988 phosphorus ban and more stringent point source TP limits established in the 1980s led to a step decrease in phosphorus concentrations. The USGS also reported a step decrease
in TP concentrations.

The TP load for this period between 1/89 and 12/99 was calculated using six estimation methods (Table 1A). The lowest CV, 0.082, was obtained using Method 6 (REG-3).

The TP load was also calculated for the data from January 1, 1990 to December 31, 1999 to determine if 1990 was a better start date than 1989 (Table 1B). The lowest CV for the load
calculation was 0.084, using Method 6.

Since the use of the data from 1/89 – 12/99 resulted in a lower CV and includes more data, this period will be used in the subsequent analysis.

The mean and standard deviation of the six methods calculating the loads for the period from 1/1/89 – 12/31/99 are shown in Figure 1.

Concentration versus Date
Figure 2 shows TP concentration versus date for the entire period (1/1/89-12/31/99). This figure illustrates the average concentration is relatively constant over the 11-year period.
Table 2 provides a statistical summary of the relationship between TP concentration and date. Figure 3 illustrates the residuals versus date for 1/1/89 – 12/31/99. There is a slight
downward slope to the regression residuals during this period.

Concentration versus Flow
Figure 4 illustrates TP concentration versus flow, using Method 6 and data from 1/89 – 12/99, and Table 3 provides the statistics associated with this graph. Figure 5 illustrates the
residuals versus flow. The relationship appears to be nonlinear with a change in pattern around the 1000 hm3/yr flow level, suggesting that better results may be achieved through flow
stratification.

Date Stratification
The data were stratified according to date to determine if there was any detectable trend in TP concentration over time since 1988. The first stratum included data from January 1, 1989
– December 31, 1994 while the second stratum included data from January 1, 1995 – December 30, 1999. The load calculations for the six estimation methods are shown in Table 4. Method
6 had the lowest CV, 0.081. Figure 6 illustrates concentration versus flow for this stratification. Figure 6 shows a decreasing relationship between TP and flow for the second stratum
(95-99). This interpretation is uncertain because there have not been very many high flow TP measurements during this period.

Seasonal Stratification
The data (1989-1999) were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method
6, 0.081, was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for the seasonal stratification using Method 6. Figure 7 suggests there is no clear
relationship between total phosphorus concentration and season.

Flow Stratification
Various flow values were tested as cutoffs for the flow stratification scheme. In each case, Method 6 had the lowest CV. The coefficients of variations for the various cutoff values
are shown below.

Flow (hm3/yr) CV # of Measurements in Stratum 1 # of Measurements in Stratum 2 200 0.078 18 116 300 0.078 35 99 400 0.078 46 88 500 0.079 58 76 600 0.080 67 67 700 0.082 76 58 800 0.085 81 53
900 0.085 86 48 1,000 0.085 87 47
The load calculations for the cutoff value that produced the lowest CV, 400 hm3/yr, are shown in Table 6. A plot of concentration versus flow for this stratification is shown in Figure
8.

Although the stratification at 1,000 hm3/yr had a higher CV than the stratification at 400 hm3/yr and even a higher CV than no stratification, visually it appears to represent the data
better than any other stratification scheme. The load calculations for stratification with 1,000 hm3/yr as the cutoff are contained in Table 7. Flow versus concentration is illustrated
in Figure 9.

Conclusions
There was a step decrease in TP concentration that corresponded to the 1988 phosphate ban and other nutrient management initiatives.
In the period from 1/89 – 12/99, there was no clear relationship between TP and either date or season.
The relationship between TP concentration and flow is nonlinear. TP concentrations decreased with flow at low stream flows, and increased with flow at high stream flows.
All of the graphs illustrated the need for stratification by flow. Although the stratification at 400 hm3/yr had the lowest CV, the stratification at 1,000 hm3/yr appeared to best represent
the data. In some respects, this criterion is more important than CV because the CV can be influenced by outliers.

Recommended Method and Stratification
Based on the conclusions and findings above, the recommended procedure for estimating TP loads for the Haw River near Bynum is:
Use Method 6 for the period 1/1/89 – 12/31/99 with flow stratification at 1000 hm3/yr

Load Calculations
Table 8 shows the annual loads for total phosphorus from 1/89 to 12/99 with Method 6 and no stratification. Table 9 shows the annual loads for total phosphorus from 1/89 to 12/99 with
Method 6 and flow stratification at 400 hm3/yr. Table 10 shows the annual total phosphorus loads obtained by using the recommended procedure: Method 6, flow stratification at 1000
hm3/yr, and data from 1/89 – 12/99. Note that FLUX provides both direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator
corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges not accounted for by date stratification.

Table 1A: Calculation of Total Phosphorus (TP) Load using Six Estimation Methods (1989-1990)

Table 1B: Calculation of TP Load using Six Estimation Methods (1990-1999)

Figure 1: Mean and one Standard Error for Six Estimation Methods
Figure 2: TP Concentration (µg/l) versus Date

Table 2: Statistics for TP Concentration versus Date
Figure 3: TP Residuals versus Date
Figure 4: TP Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for TP Concentration versus Flow

Figure 5: TP Residuals versus Flow

Table 4: TP Load Calculations for Date Stratification (1989-1994 and 1995-1999)
Figure 6: TP Concentration (µg/l) versus Flow (hm3/yr) For Date Stratification
Table 5: TP Load Calculations with Seasonal Stratification

Figure 7: TP Concentration (µg/l) versus Flow (hm3/yr) for Seasonal Stratification

Table 6: TP Load Calculations for Flow Stratification at 400 hm3/yr
Figure 8: TP Concentration (µg/l) vs Flow (hm3/yr) with Stratification at 400 hm3/yr
Table 7: TP Load Calculations with Flow Stratification at 1000 hm3/yr

Figure 9: TP Concentration (µg/l) versus Flow (hm3/yr) with Stratification at 1000 hm3/yr
Table 8: Annual TP Loads with No Stratification

Table 9: Annual TP Loads with Flow Stratification at 400 hm3/yr

Table 10: Annual TP Loads with Flow Stratification at 1000 hm3/yr
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.

18

7

Appendix D/Morgan FLUX.doc

FLUX Analysis of Nutrient Loading
on Morgan Creek near Farrington Road
January 1, 1984 – October 31, 1999

September 15, 2000

Introduction

Lake water quality models require complete or continuous estimates of tributary nutrient loads. Observed nutrient data from tributaries consists of point-in-time measurements of concentration,
along with continuous estimates of flow from USGS gaging stations. Converting such point-in-time concentrations to continuous load estimates presents a number of technical challenges
because concentration and flow are typically correlated instead of being independent of one another. A variety of methods are available to address this problem, but no one method is
superior for all situations. The US Army Corps of Engineers Waterways Experiment Station produced the FLUX program to evaluate and compare multiple methods of estimating nutrient loads
from tributary concentration data. Tetra Tech used FLUX to analyze nutrient loads for each of the monitored tributaries to Jordan Lake.

The FLUX analysis required input files containing mean daily flows and sample data consisting of measured concentrations of various nutrient species. The daily mean flows of the Morgan
Creek were obtained from the USGS and the water quality monitoring data were a combination North Carolina Department of Water Quality and OWASA WRF data. There was nutrient data available
for one day from the USGS and it was included as well. One OWASA measurement for NH3 (3/18/98) was excluded from the analysis. The measured concentration of NH3 on that day was 64.4
mg/l, a value significantly higher than any other recorded measurement.

The data analysis was conducted through October 1999, the month through which monitoring data were available. The start date of the analysis was set to January 1, 1984. Although the
new Y2K-compliant version of FLUX can handle more than 6000 flow values, 1984 is a little more than 2 years after Jordan Lake was impounded. For several nutrient species, the analysis
indicated it was better to use and even later start date.

Table 1 below compares the entire flow record from January 1, 1984 – October 31, 1999 to the flow record for the dates on which nutrient measurements were made (sample flows).

Number of Measurements Flow Range
(cfs) Mean Value
(cfs) Median Value
(cfs) Entire Flow Record 187 6.5-2600 44 18 Flow on days when nutrient measurements were taken
5782 3-349 32 17
Table 1 indicates that the sample flow values have a median value and mean value very close to those of the entire flow record. However, the sample values do not include values from
the high end of the flow range. It is possible for the sample flows to be lower than the average daily flow because many of the sample flows were instantaneous flow values.

Table of Contents
Introduction 2
Table of Contents 3
Total Nitrogen 4
Nitrate/Nitrite 15
Total Kejeld Nitrogen 28
Ammonia 41
Total Phosphorus 52

Total Nitrogen
FLUX analyzed 187 total nitrogen samples and 5782 daily flow values from January 1, 1984 through October 31, 1999. The total nitrogen load for this period was calculated using six estimation
methods (Table 1A). Method 6 was used in the subsequent analysis because it has the lowest coefficient of variation (CV), .042, of all the methods. The load calculations were also
performed with an abridged data record from 1/1990 – 10/1999 (Table 1B). The lowest CV for these calculations was .058 for Method 6. Since the CV was higher for the smaller record
and Figure 2 (below) shows no clear trend with time, the entire record (1/84 – 10/99) was used in the subsequent analysis. Figure 1 shows the mean and standard deviations of the six
methods.

Concentration versus Date
Figure 2 shows total nitrogen concentration versus date for the entire period. Table 2 provides a statistical summary of the relationship between nitrogen concentration and date. Figure
3 shows the residuals versus date; the residuals gradually decrease with date. Figures 2 and 3 illustrate that total nitrogen concentrations in Morgan Creek have remained relatively
constant between 1984 and 1999.

Concentration versus Flow
Figure 4 illustrates nitrogen concentration versus flow using Method 6 and Table 3 provides the statistics associated with this graph. Figure 5 illustrates the residuals versus flow.
Figure 4 illustrates that the total nitrogen concentration decreases as flow increases.

Date Stratification
The data were stratified according to date to determine if there was a pronounced difference in nitrogen concentrations in the two halves of the 15.8-year record. The first stratum
included data from January 1, 1984 – December 31, 1992 while the second stratum included data from January 1, 1993 – October 31, 1999. The load calculations for the six estimation
methods are show in Table 4. Method 6 had the lowest CV, .042. Figure 6 shows a plot of concentration versus flow for this stratification. Figure 6 shows there is no trend in nitrogen
concentration with time because the two trendlines cross in the middle of the data.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 6, .041,
was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for the seasonal stratification. The graph illustrates the relationship between total nitrogen
concentration and flow is almost identical for the growing (May – September) and non-growing (October-April) seasons.

Flow Stratification
For the stratification of flow at 40 cubic hectometers/year, the lowest CV was .040 (Method 4) as shown in Table 6. Figure 8 shows a plot of concentration versus flow for stratification
at 40. Figure 8 illustrates the relationship between total nitrogen concentration and flow is almost identical for low and high flows.

Conclusions
Method 6 was best for no stratification and date stratification while Method 4 was best for seasonal and flow stratification.
The plot of nitrogen concentration versus time and the date stratification indicate that total nitrogen concentrations have remained relatively constant on Morgan Creek from 1984 to
1999.
There was no relationship between total nitrogen concentration and season.
The plot of nitrogen concentration versus flow and the flow stratification indicated that concentration decreases as flow increases and this relationship is the same for low and high
flows.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for total nitrogen calculations on Morgan Creek is:
Use Method 6 for the entire period (1/84 – 10/99) with no stratification

Load Calculations
Table 7 shows the annual total nitrogen loads for the entire period using Method 6 and no stratification using a maximum interpolation interval of 12 days. Note that FLUX provides both
direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as
might be caused by changes in WWTP discharges not accounted for by date stratification. Figure 9 illustrates the time series estimates for the load calculations in Table 7. Also note
that results from 1999 reflect only a partial year, and are therefore not directly comparable to annual totals from previous full years.

Table 1A: Calculation of Total Nitrogen Loads on Morgan Creek using Six Estimation Methods (84-99)
Table 1B: Calculation of Total Nitrogen Loads on Morgan Creek using Six Estimation Methods (90-99)

Figure 1: Mean and one Standard Error for Six Estimation Methods
Figure 2: Total Nitrogen Concentration versus Date

Table 2: Statistics for Graph of Total Nitrogen Concentration versus Date

Figure 3: Plot of Residuals versus Date

Figure 4: Nitrogen Concentration versus Flow
Table 3: Statistics of Nitrogen Concentration versus Flow
Figure 5: Residuals versus Flow
Table 4: Calculation of Nitrogen Loads with Six Estimation Methods using Date Stratification
Figure 6: Nitrogen Concentration versus Flow for Date Stratification
(1:84-92, 2:93-99)
Table 5: Load Calculations for Total Nitrogen using Season Stratification

Figure 7: Total Nitrogen Concentration versus Flow for Season Stratification
Strat 1= Oct-April; Strat 2= May-September
Table 6: Load Calculations for Flow Stratification at 40 hm3/yr

Figure 8: Concentration versus Flow for Flow Stratification at 40 hm3/yr

Table 7: Annual Total Nitrogen Loads (Method 6 and no stratification)

Figure 9: Time Series Load Estimates

Nitrate/Nitrite

FLUX analyzed 184 nitrate/nitrite samples and 5783 daily flow values from January 1, 1984 through October 31, 1999. The nitrate/nitrite load for this period was calculated using six
estimation methods (Table 1A). Method 6 was used in the subsequent analysis because it had the lowest coefficient of variation (CV), .052, of all the methods. The load calculations
were also performed with an abridged data record from 1/1990 – 10/1999 (Table 1B). The lowest CV for these calculations was .069 for Method 1. Since the CV was higher for the smaller
record and Figure 2 (below) shows no clear trend with time, the entire record (1/84 – 10/99) was used in the subsequent analysis. Figure 1 shows the mean and standard deviations of
the six methods.

Concentration versus Date
Figure 2 shows nitrate/nitrite concentration versus date for the entire period. Table 2 provides a statistical summary of the relationship between nitrate/nitrite concentration and
date. Figure 3 shows the residuals versus date; the residuals gradually decrease with date. Figures 2 and 3 illustrate that nitrate/nitrite concentrations in Morgan Creek have remained
relatively constant between 1984 and 1999.

Concentration versus Flow
Figure 4 illustrates nitrate/nitrite concentration versus flow using Method 6 and Table 3 provides the statistics associated with this graph. Figure 5 illustrates the residuals versus
flow. Figure 4 illustrates that the nitrate/nitrite concentration decreases as flow increases.

Date Stratification
The data were stratified according to date to determine if there was a pronounced difference in nitrate/nitrite concentrations in the two halves of the 15.8-year record. The first stratum
included data from January 1, 1984 – December 31, 1991 while the second stratum included data from January 1, 1992 – October 31, 1999. The load calculations for the six estimation
methods are show in Table 4. Method 6 had the lowest CV, .056. Figure 6 shows a plot of concentration versus flow for this stratification. Figure 6 shows there is no trend in nitrate/nitrite
concentration with time because the two trend lines cross in the middle of the data.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 6, .051,
was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for the seasonal stratification. The graph illustrates the relationship between total nitrogen
concentration and flow is almost identical for the growing (May – September) and non-growing (October-April) seasons.

Flow Stratification
For the stratification of flow at 40 cubic hectometers/year, the lowest CV was .048 (Method 6) as shown in Table 6. Figure 8 shows a plot of concentration versus flow for stratification
at 40. Figure 8 illustrates a trend line that decreases for low flows and a trend line that decreases more rapidly for high flows indicating that nitrate/nitrite concentrations are
stratified by flow. Figure 9 shows a plot of residuals versus time and Figure 10 shows a plot of residuals versus flow for the stratified data regression. Table 7 and Table 8 provide
the statistics for Figures 9 and 10, respectively.

Conclusions
Method 6 was the best for calculating NOx loads for all stratification schemes.
The plot of nitrate/nitrite concentration versus time and the date stratification indicate that nitrate/nitrite concentrations have remained relatively constant on Morgan Creek from
1984 to 1999.
There was no relationship between nitrate/nitrite concentration and season.
The plot of nitrate/nitrite concentration versus flow and the flow stratification indicated that concentration decreases as flow increases and this relationship is more pronounced for
high flows than for low flows. The difference in trends between high and low flows is significant and stratification by flow will produce the most accurate nitrate/nitrite load estimates.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for nitrate/nitrite calculations on Morgan Creek is:
Use Method 6 for the entire period (1/84 – 10/99) with flow stratification.

Load Calculations
Table 9 shows the annual total nitrogen loads for the entire period using Method 6 with flow stratification using a maximum interpolation interval of 12 days. Note that FLUX provides
both direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such
as might be caused by changes in WWTP discharges not accounted for by date stratification. Figure 11 illustrates the time series estimates for the load calculations in Table 9. Also
note that results from 1999 reflect only a partial year, and are therefore not directly comparable to annual totals from previous full years.

Table 1A: Calculation of NOx Load on Morgan Creek using Six Methods (84-99)

Table 1B: Calculation of NOx Load on Morgan Creek using Six Methods (90-99)

Figure 1: Mean one Standard Error for Six Load Estimation Methods

Figure 2: Total NOx Concentration versus time (84-99)

Table 2: Statistics for Figure 2

Figure 3: Residuals versus Date (84-99)

Figure 4: NOx Concentration versus Flow
Table 3: Statistics for Figure 4

Figure 5: Residuals versus Flow

Table 4: Load Calculation with Date Stratification (84-91, 92-99)

Figure 6: NOx concentration versus Flow for Date Stratification

Table 5: Load Calculation with Season Stratification

Figure 7: NOx versus Flow for Season Stratification
(Strat 1 = Oct-April; Strat 2=May-Sept)

Table 6: Load Calculations for Flow Stratification at 40 hm3/yr

Figure 8: Concentration versus Flow for Stratification at 40 hm3/yr

Figure 9: Residuals versus Date for Stratification at 40 hm3/yr

Table 7: Statistics for Figure 9

Figure 10: Residuals versus Flow for Stratification at 40 hm3/yr

Table 8: Statistics for Figure 10

Table 9: Annual NOx Loads with Flow Stratification at 40 hm3/yr

Figure 11: Time Series Load Estimates

Total Kjeldahl Nitrogen

FLUX analyzed 187 total Kjeldahl nitrogen (TKN) samples and 5783 daily flow values from January 1, 1984 through October 31, 1999. The TKN load for this period was calculated using six
estimation methods (Table 1A). Method 4 was used in the subsequent analysis because it had the lowest coefficient of variation (CV), .034, of all the methods. Figure 1 shows the mean
and standard deviations of the six methods.

Concentration versus Date
Figure 2 shows TKN concentration versus date for the entire period (1/84-10/99) using Method 4 and Table 2 provides a statistical summary of the relationship between TKN concentration
and date. Figure 2 illustrates the TKN concentration was relatively constant over the 16-year period except for some occasionally extreme values.

Figure 3 illustrates the residuals versus date for the period between 1/84 – 10/99. The residuals have a slight downward slope for this period. However, this downward trend may result
from the extreme high values reported in the 1984-1986 period.

Concentration versus Flow
Figure 4 illustrates TKN concentration versus flow using Method 4 and Table 3 provides the statistics associated with this graph. Figure 4 illustrates that TKN concentrations decrease
slightly as flow increases. Figure 5 illustrates the residuals versus flow and Table 4 provides the statistics for the residuals plot. The residuals are constant with respect to changing
flows, indicating that Method 4 is a good predictor of TKN concentrations at high and low flows.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in TKN concentrations with time. The first stratum contained data from 1/84 – 12/91 while
the second stratum contained data from 1/92 – 10/99. Method 4 had the lowest CV for this stratification, .033 (Table 5). Figure 6 shows concentration versus flow using Method 4.
Figure 6 shows the trendlines for the two strata crossing in the middle of the data. Thus, there is no clear trend between TKN concentration and date.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 6 illustrates the load calculations for each estimation method. The CV for Method 4 was .040,
the lowest of all the methods. Figure 7 illustrates concentration versus flow for the season stratification. Figure 7 illustrates that TKN concentrations are relatively constant across
changing flows during the growing season, but decrease as flows increase during the non-growing “winter” season. However, the data does not support seasonal stratification of the data
to estimate annual TKN loads.

Flow Stratification
For the stratification of flows greater than and less than 10 cubic hectometers/year, the lowest CV was .034 for Methods 2, 3, and 4 (Table 7). Because Method 4 is designed to account
for a relationship between flow and concentration, it was used to plot concentration versus flow (Figure 8). Figure 8 illustrate TKN concentrations increase with flow for low flows
and then decrease with flow for high flows. This stratification, however, is not robust because it is heavily reliant on extreme events. Figure 9 shows the plot of residuals versus
flow for the flow stratified Method 4 regression and Table 8 provides the statistics for Figure 9. The residuals are nearly constant across the full range of flow values.

Conclusions
There was no clear relationship between TKN concentration and date (a trend consistent with USGS findings for the period from 1985 to 1995).
The stratification by seasons indicated that TKN concentrations were high during the growing season during high flows. The CV for this stratification was the lowest obtained in the
TKN analysis.
Although there may be a change in the relationship between TKN concentrations and flow between low and high flows there is no improvement accuracy of load estimates is obtained by stratifying
data according to flow (see Figures 10 and 11 regarding Load Calculations).

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for TKN calculations for Morgan Creek is:
Use Method 4 for the period 1/1/84 – 10/31/99 with no stratification

Load Calculations
Table 9 shows the annual loads for TKN with Method 4 and flow stratification at 10 hm3/yr using a maximum interpolation interval of 12 days. Note that FLUX provides both direct ‘model’
outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused
by changes in WWTP discharges not accounted for by date stratification. Figure 10 illustrates the time series of calculated loads from Table 9. Also note that results from 1999 reflect
only a partial year, and are therefore not directly comparable to annual totals from previous full years. Table 10 shows the annual TKN loads obtained by using a maximum interpolation
interval of 12 days and Method 4 with no stratification. Figure 11 shows the time series of load estimates from Table 10.

Table 1: Calculation of TKN Load on Morgan Creek using Six Methods (84-99)

Figure 1: Mean and one Standard Error for Six Estimation Methods (84-99)

Figure 2: TKN Concentration versus Date (84-99)

Table 2: Statistics for Figure 2

Figure 3: Residuals versus Date

Figure 4: TKN Concentration versus Flow

Table 3: Statistics for Figure 4 Figure

Figure 5: Residuals versus Flow

Table 4. Statistics for Figure 5

Table 5: Load Calculations for Date Stratification
(Strata 1 = 1984-91; Strata 2 = 1992-1999)

Figure 6: Flow versus Concentration for Date Stratification

Table 6: Load Calculations for Seasonal Stratification
(Strata 1=Oct-April; Strata 2=May-Sept)

Figure 7: Concentration versus flow for Seasonal Stratification

Table 7: Load Calculations for Flow Stratification (0-10; >10 hm3/yr)

Figure 8: Concentration versus flow for Flow Stratification (0-10; >10 hm3/yr)

Figure 9: Residuals versus Flow with Flow Stratification at 10 hm3/yr

Table 8: Statistics for Figure 9

Table 9: Annual TKN loads with Method 4 and no Stratification

Figure 10: Time Series Load Estimates for Method 4 and no Stratification

Table 10: Annual TKN loads with Method 4 with Flow Stratification

Figure 11: Time Series Load Estimates for Method 4 with Stratification

Ammonia
FLUX analyzed 184 ammonia (NH3) samples from January 1, 1984 through October 31, 1999. The NH3 load for this period was calculated using six estimation methods (Table 1). The lowest
CV, .092, was obtained using Method 5. The mean and standard deviation of the six methods calculating loads with data from 1/84 – 10/99 are shown in Figure 1.

Concentration versus Date
Figure 2 shows NH3 concentration versus date for the entire period (1/84 – 10/99) using Method 5 and Table 2 provides a statistical summary of the relationship between NH3 concentration
and date. Figure 2 illustrates that ammonia measurements were relatively constant over the period examined.

Figure 3 illustrates the residuals versus date for the period between 1/87 and 10/99 using Method 5 and Table 3 provides the statistics for Figure 3. There is a slight downward slope
to the regression residuals during this period suggesting the potential for a slight decrease in ammonia concentration with time.

Concentration versus Flow
Figure 4 illustrates NH3 concentration versus flow and Table 4 provides the statistics associated with this graph. Figure 4 indicates that NH3 concentrations increase slightly as flow
increases, but the slope significance statistic provided in Table 4 indicates that the relationship between the two variables is very weak. Figure 5 illustrates the residuals versus
flow and Table 5 provides the associated statistics. The residuals remain constant as flow increases and indicate that the regression method exhibits consistent error in predicting
concentrations.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in NH3 concentration over time. The first stratum included data from January 1, 1984 – December
31, 1991 while the second stratum included data from January 1, 1992 – October 31, 1999. The load calculations for the six estimation methods are shown in Table 6, with the lowest
CV, .121, associated with Method 4. Figure 6 illustrates concentration versus flow using Method 4 for this date stratification. Figure 6 shows there is no trend in NH3 concentration
with time because the two trend lines cross in the middle of the data.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 7 illustrates the load calculations for each estimation method. The CV for Methods 4 and
6 were the lowest of the estimation methods at .130 while the CV for Method 2 was .134. Figure 7 illustrates concentration versus flow for the season stratification using Method 4.
Method 4 was utilized over Method 6 because Method 6 relies on a strong relationship between concentration and flow, which is not present in this case (see Figure 4 and Table 4).
Figure 7 illustrates that NH3 concentrations are slightly lower in the non-growing season (October – April) than in the growing season (May – September)and that concentrations increase
as flow increases in the non-growing season while remaining relatively constant across the range of flows in the growing season. Seasonal stratification appears to be significant in
this analysis.

Flow Stratification
For the stratification of flows greater than and less than 20 cubic hectometers/year, the lowest CV was .143 for Method 2 (Table 8). This stratification was used to plot concentration
versus flow (Figure 8). Figure 8 illustrates that NH3 concentrations are slightly higher for the upper flow strata (above 20 hm3/yr.), but the difference is not pronounced.

Conclusions
Method 5 was the best for determining load calculations for the data from 1/84 – 10/99.
Date stratification showed no difference in NH3 concentrations between the 1/84 – 12 – 91 strata and the 1/92 - 10/99 strata.
The seasonal stratification showed a pronounced difference in NH3 concentrations and a significant difference in concentration trends with regards to flow between the growing and non-growing
seasons.
Flow stratification does not capture the fluctuation of data as well as seasonal stratification.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for NH3 calculations on the Morgan Creek is:
Use Method 4 for the period 1/1/84 – 10/31/1999 with seasonal stratification.

Load Calculations
Table 10 shows the annual loads for NH3 using Method 4 and seasonal stratification using a maximum interpolation interval of 12 days. Note that FLUX provides both direct ‘model’ outputs
and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes
in WWTP discharges not accounted for by date stratification. Figure 9 shows the time series estimates for annual load calculations in Table 10. Also note that results from 1999 reflect
only a partial year, and are therefore not directly comparable to annual totals from previous full years.

Table 1: Calculation of NH3 Loads using Six Estimation Methods (84-99)

Figure 1: Mean one Standard Error for Deviation of Six Methods

Figure 2: NH3 Concentration versus Date

Table 2: Statistics for Figure 2

Figure 3: Residuals versus Date (87-99)

Table 3: Statistics for Figure 3

Figure 4: NH3 Concentration versus Flow

Table 4: Statistics for Figure 4

Figure 5: Residuals versus Flow

Table 5: Statistics for Figure 5

Table 6: Load Calculations with Date Stratification (84-91; 92-99)

Figure 6: NH3 Concentration versus Flow for Date Stratification (84-91; 92-99)

Table 7: Load Calculations with Seasonal Stratification
Strat 1= Oct – April; Strat 2 = May - September

Figure 7: NH3 Concentration versus Flow for Seasonal Stratification
Strat 1= Oct – April; Strat 2 = May – September

Table 8: Load Calculations with Flow Stratification (0-20, >20 hm3/yr)

Figure 8: Concentration versus Flow for Method 2 with Flow Stratification

Table 9: Annual NH3 loads with seasonal stratification and Method 4

Figure 9: Time Series Estimates for NH3 loads

Total Phosphorus
FLUX analyzed total phosphorus samples from January 1988 through October 1999. Although data were available from earlier years, previous FLUX analysis confirmed the 1988 phosphorus
ban led to a step decrease in phosphorus concentration in the Haw River near Bynum and a similar trend would be expected in Morgan Creek. Furthermore, the USGS also reported a decrease
in total phosphorus concentration in the Morgan Creek between 1983 and 1995. The phosphorus load for the period between 1/88 and 10/99 was calculated using six estimation methods (Table
1A). The lowest CV, .053, was obtained using Method 3.

The phosphorus load was also calculated for the data from January 1, 1989 to October 31, 1999 to determine if 1989 was a better start date than 1988 (Table 1B). The lowest CV for the
load calculation was .050, using Method 3. Since the use of the data from 1/89 – 10/99 resulted in a slightly lower CV, this period was used in the subsequent analysis, unless otherwise
noted. The mean and standard deviation of the six methods calculating the loads for the period from 1/1/89 – 10/3/99 are shown in Figure 1.

Concentration versus Date
Figure 2 shows total phosphorus concentration versus date using Method 3 for the entire period and Table 2 provides a statistical summary of the relationship between total phosphorus
concentration and date. Figure 3 illustrates the residuals versus date and Table 3 provides the statistics for this relationship. Observed data plotted in Figure 2 illustrate that
there were more measurements of low concentrations in the 1990s than in the 1980, and Figure 3 shows the residuals decreasing with time, suggesting that the phosphorus concentration
has decreased with time.

Concentration versus Flow
Figure 4 illustrates phosphorus concentration versus flow using Method 3 and Table 4 provides the statistics associated with this graph. Figure 5 illustrates the residuals versus flow;
the residuals decrease as flow increases, indicating that Method 3 results in a degree of bias in load calculations at high and low flow. Table 5 provides the statistics for Figure
5.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in total phosphorus concentration over time. The first stratum included data from January
1, 1989 – December 31, 1993 while the second stratum included data from January 1, 1994 – October 31, 1999. The load calculations for the six estimation methods are shown in Table
6. Method 2 had the lowest CV, .058. Figure 6 illustrates concentration versus flow for this stratification using Method 2 and Table 7 contains the statistics for this relationship.
Figure 6 shows phosphorus concentrations decreased between the two strata. However, the CV values associated with the calculation methods increased for almost all methods relative
to the unstratified data indicating that 1993/1994 may not be the best break point to define the strata.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 6 illustrates the load calculations for each estimation method. The CV for Method 2, 0.071,
was the lowest of the estimation methods, and Method 4 also resulted in a relatively low CV of 0.072. Figure 7 illustrates concentration versus flow for the seasonal stratification
using Method 2 and Figure 8 illustrates concentration versus flow for the seasonal stratification using Method 4. Concentrations were plotted using Method 4 to visually evaluate the
suitability of the best regression method for this stratification. Figures 7 and 8 illustrate the total phosphorus concentrations were higher in the non-growing season (October – April).

Flow Stratification
For the flow stratification at 40 cubic hectometers/year, the lowest CV was .054 for Method 3. Table 9 shows the load calculations for this stratification. A plot of concentration
versus flow for this stratification is shown in Figure 9. Figure 9 shows that phosphorus concentrations are lower for the upper flow strata. In addition, Figure 10 shows a plot of
concentration vs. flow for the flow-stratified data using Method 4 to calculate loads. In Figure 10 the regression method (Method 4) estimates indicate that phosphorus concentrations
decrease with increasing flows in the lower flow strata and increase with increasing flow in the higher strata.

Flow and Date Stratification
Due to the pronounced results of flow and date stratification analyses, data were broken into four strata on the basis of flow and date as follows:

STRATA 1: January 1988 – September 1991, Flows < 40 cubic hectometers/year
STRATA 2: January 1988 – September 1991, Flows => 40 cubic hectometers/year
STRATA 1: October 1991 – October 1999, Flows < 40 cubic hectometers/year
STRATA 1: October 1991 – October 1991, Flows => 40 cubic hectometers/year

Data were date stratified at October 1991 in order to reflect completion of the upgrade to biological nutrient removal at the OWASA Mason Farm WWTP, which dominates flow and nutrient
loads in Morgan Creek. Table 10 shows the load calculations for this stratification with Method 4 having the lowest CV at 0.54. A plot of concentration versus date for this stratification
is shown in Figure 11 which shows a noticeable decrease in concentrations after the WWTP upgrade . A plot of concentration versus flow for this stratification is shown in Figure 12,
which shows a pronounced difference in concentrations between the two time periods within the lower flow strata and and inverse relationships of concentration to flow for the higher
flow strata.

Conclusions
Method 3 resulted in the lowest CVs for load calculations for all stratification schemes except seasonal and date stratification.
In the period from 1/88 – 10/99, there was a significant decrease in phosphorus concentrations. The decrease most likely resulted from the upgrade of the OWASA – Mason Farm WWTP completed
in fall of 1991.
There was a slight relationship between phosphorus concentration and season which may be a function of the relationship to flow.
The flow stratification illustrates that the relationship between phosphorus concentration varies significantly across flow regimes in Morgan Creek.
Best fit to observed data was achieved when data where stratified by date and flow.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for total phosphorus concentrations on the Morgan Creek is:
Use Method 4 for the period 1/88-10/99 with flow and date stratification

Load Calculations
Table 11 shows the annual loads for total phosphorus from 1/88 to 10/99 with Method 4 and flow and date stratification (as outlined above) using a maximum interpolation interval of 12
days. Note that FLUX provides both direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial
correlation in results, such as might be caused by changes in WWTP discharges not accounted for by date stratification. Figure 13 illustrates the time series load estimates for the
load calculations in Table 11. Also note that results from 1999 reflect only a partial year, and are therefore not directly comparable to annual totals from previous full years.

Table 1A: Calculation of Total Phosphorus Load using Six Methods (88-99)

Table 1B: Calculation of Total Phosphorus Load using Six Methods (89-99)

Figure 1: Mean and Standard Deviation of Six Estimation Methods

Figure 2: Total Phosphorus Concentration versus Date

Table 2: Statistics for Figure 2

Figure 3: Residuals versus Date

Table 3: Statistics for Figure 3

Figure 4: Total Phosphorus Concentration versus Flow

Table 4: Statistics for Figure 4

Figure 5: Residuals versus Flow

Table 5: Statistics for Figure 5

Table 6: Load Calculations for Date Stratification (89-93, 94-99)

Figure 6: Concentration versus flow For Date Stratification (89-93, 94-99)

Table 7: Statistics for Figure 6

Table 8: Load Calculations with Seasonal Stratification

Figure 7: TP Concentration versus Flow for Seasonal Stratification (Method 2)

Figure 8: TP Concentration versus Flow for Seasonal Stratification (Method 4)

Table 9: Load Calculations for Flow Stratification at 40 hm3/yr

Figure 9: Total Phosphorus Concentration versus Flow for Flow Stratification at 40 hm3/yr (Method 3)

Figure 10: Total Phosphorus Conc. vs Flow for Flow Stratification at 40 hm3/yr (Method 4)

Table 10: Load Calculations for Flow Stratification at 40 hm3/yr and Date Stratification (Date Strat = Jan 1988 – Sept 1991; Oct 1991- Oct 1999)
Figure 11: Total Phosphorus Conc. vs Time for Flow & Date Strat. (Method 4)

Figure 12: Total Phosphorus Conc. vs Flow for Flow & Date Strat. (Method 4)
Table 11: Time Series Load Estimates from Method 4, with Flow Stratification at 40 hm3/yr and Date Stratification (Jan 1988 – Sept 1991; Oct 1991- Oct 1999)

Figure 13: Time Series Load Estimates from Method 4 with Flow & Date Stratification

U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.

3

Appendix D/New Hope FLUX.doc

FLUX Analysis of Nutrient Loading
on the New Hope Creek near Blands
January 1, 1984 – April 30, 2000

April 26, 2001

This document has been prepared by Tetra Tech, Inc. for the
Jordan Lake Nutrient Response Modeling Project.
Funding for this project is provided by:

City of Burlington
City of Graham
City of Greensboro
Town of Mebane
Orange Water and Sewer Authority
Town of Pittsboro
City of Reidsville

Table of Contents
Table of Contents 2
Introduction 3
Total Nitrogen 4
Nitrate/Nitrite 15
Total Kjeldhal Nitrogen 26
Ammonia 37
Total Phosphorus 48

Introduction

To properly calibrate and verify the Jordan Lake Nutrient Response Model, complete or continuous estimates of tributary nutrient loads must be available or developed. Observed nutrient
data from tributaries consists of point-in-time measurements of concentration, along with continuous estimates of flow from USGS gaging stations. Converting such point-in-time concentrations
to continuous load estimates presents a number of technical challenges because concentration and flow are typically correlated instead of being independent of one another. A variety
of methods are available to address this problem, but no one method is superior for all situations. The US Army Corps of Engineers Waterways Experiment Station produced the FLUX program
to evaluate and compare multiple methods of estimating nutrient loads from tributary concentration data. Tetra Tech used FLUX to analyze nutrient loads for each of the four monitored
tributaries to Jordan Lake. These tributaries are (a) the Haw River; (b) New Hope Creek; (c) Morgan Creek; and (d) Northeast Creek. This report presents the results of the FLUX analysis
for New Hope Creek near Blands.

The FLUX analysis required input files containing mean daily flows and sample data consisting of measured concentrations of various nutrient species. The daily mean flows of the New
Hope Creek were obtained from the USGS (gage 02097314) and the water quality monitoring data were a combination of USGS data collected at the flow gage, North Carolina Division of Water
Quality, and South Durham Water Reclamation Facility data. The flow gage is located on the right bank, 15 feet downstream of Secondary Road 1107 in Durham County. Water quality data
from all three sources were collected at the SR 1107 bridge crossing.

The data analysis was conducted through April 2000, the month through which monitoring data were available. The start date of the analysis was set to January 1, 1984. Although the
new Y2K-compliant version of FLUX can handle more than 6000 flow values, 1984 is a little more than 2 years after Jordan Lake was impounded. For several nutrient species, the analysis
indicated it was better to use an even later start date.

Table 1 below compares the entire flow record from January 1, 1984 – April 30, 2000 to the flow record for the dates on which nutrient measurements were made (sample flows).

Table 1. Flow Data Comparison
Number of Measurements Flow Range
(cfs) Mean Value
(cfs) Median Value
(cfs) Entire Flow Record 5,965 .39 – 6,300 102 35 Flow on days when nutrient measurements were taken 316 5 – 3,420 104 33
Table 1 indicates that the sample flow values have a median value and mean value very close to those of the entire flow record. However, the sample values do not include values from
the high end of the flow range.

Total Nitrogen

FLUX was used to analyze 310 total nitrogen (TN) samples and 5,965 daily flow values from January 1, 1984 through April 30, 2000. The TN load for this period was calculated using six
estimation methods (Table 1A). Method 4 (REG-1) was used in the subsequent analysis because it has the lowest coefficient of variation (CV), 0.033, of all the methods. The load calculations
were also performed with an abridged data record from 1/1990 – 4/2000 (Table 1B). The lowest CV for these calculations was 0.040 for Method 4 (REG-1). Since the CV was higher for
the smaller record and Figure 2 (below) shows no clear trend with time, the entire record (1/84 – 4/00) was used in the subsequent analysis. Figure 1 shows the mean and standard deviations
of the six methods.

Concentration versus Date
Figure 2 shows TN concentration versus date for the entire period. Table 2 provides a statistical summary of the relationship between TN concentration and date. Figure 3 shows the
residuals versus date; the residuals gradually decrease with date. Figures 2 and 3 illustrate that TN concentrations in the New Hope Creek have remained relatively constant between
1984 and 2000.

Concentration versus Flow
Figure 4 illustrates TN concentration versus flow using Method 4 and Table 3 provides the statistics associated with this graph. Figure 4 demonstrates that the TN concentrations decrease
as flow increases, but the relationship does not appear be consistent for the highest flow events. Figure 5 illustrates the residuals versus flow.

Date Stratification
The data were stratified according to date to determine if there was a pronounced difference in TN concentrations in the two halves of the 16.3-year record. The first stratum included
data from January 1, 1984 – December 31, 1992 while the second stratum included data from January 1, 1993 – April 30, 2000. The load calculations for the six estimation methods are
show in Table 4. Method 4 had the lowest CV, 0.029. Figure 6 shows a plot of concentration versus flow for this stratification. Figure 6 indicates that TN concentrations were slightly
higher for the period between 1984 – 1992. Since both strata show trendlines with similar slopes and Figure 2 showed no significant change in flow-normalized TN concentration with
time, there is no need for date stratification.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 4, 0.041,
was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for the seasonal stratification. The graph illustrates the relationship between TN concentration
and flow is almost identical for the growing (May – September) and non-growing (October-April) seasons.

Flow Stratification
For the stratification of flows at 100 cubic hectometers/year (hm3/yr), the lowest CV was 0.034 (Method 4) as show in Table 6. Figure 8 shows a plot of concentration versus flow for
stratification at 100 hm3/yr. Figure 8 illustrates the relationship between TN concentration and flow in the New Hope Creek changes. For low flows, TN concentrations decrease steadily
with increasing flows. However, for flows greater than 100 hm3/yr, TN concentrations decrease much more gradually as flow increases.

Conclusions
Method 4 was the best for all stratification schemes (and no stratification).
TN concentration appears to have decreased slightly over time between 1984 and 2000. However, the USGS reports no statistically significant trends in TN concentration at New Hope Creek
near Blands between 1983 and 1995 using the non-parametric seasonal Kendall test.
There was no relationship between TN concentration and season.
Flow stratification indicated that TN concentrations decrease steadily with increasing flows for low flows but decrease very gradually with increasing flows for high flows.

Recommended Method and Stratification
Based on the conclusions and findings above, the recommended procedure for estimating TN loads for New Hope Creek near Blands is:
Use Method 4 for the entire period (1/84 – 4/00) with flow stratification at 100 cubic hectometers/year

Load Calculations
Table 7 shows the annual TN loads for the entire period using Method 4 and no stratification using a maximum interpolation interval of 12 days. Table 8 shows the annual TN loads with
the recommended procedure, Method 4 and flow stratification at 100 cubic hectometers/year using a maximum interpolation interval of 12 days. Note that FLUX provides both direct ‘model’
outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused
by changes in WWTP discharges not accounted for by date stratification. Figure 9 illustrates the time series estimates for the load calculations in Table 8. Also note that year 2000
results reflect only the first third of the year, and are therefore not directly comparable to annual totals from previous years.

Table 1A: Calculation of TN Loads on New Hope Creek using Six Estimation Methods (1984-2000)
Table 1B: Calculation of TN Loads on New Hope Creek using Six Estimation Methods (1990-2000)

Figure 1: Mean and one Standard Error for Six Estimation Methods
Figure 2: TN Concentration (µg/l) versus Date

Table 2: Statistics for Graph of TN Concentration versus Date

Figure 3: Plot of TN Residuals versus Date

Figure 4: TN Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for TN Concentration versus Flow

Figure 5: TN Residuals versus Flow (hm3/yr)
Table 4: Calculation of TN Loads with Six Estimation Methods using Date Stratification (Stratum 1: 1984-1992, Stratum 2: 1993-2000)
Figure 6: TN Concentration (µg/l) versus Flow (hm3/yr) for Date Stratification
Table 5: TN Load Calculations with Seasonal Stratification

Figure 7: TN Concentration (µg/l) versus Flow (hm3/yr) for Seasonal Stratification
Strat 1= Oct-April; Strat 2= May-September

Table 6: TN Load Calculations with Flow Stratification at 100 hm3/yr

Figure 8: TN Concentration (µg/l) versus Flow with Flow Stratification at 100 hm3/yr

Table 7: Annual TN Loads (Method 4 and no stratification)

Table 8: Annual TN Loads (Method 4 with flow stratification at 100 hm3/yr)

Figure 9: Time Series of TN Load Estimates

Nitrate/Nitrite
Using FLUX, the nitrate/nitrite (NOx) loads from January 1984 – April 2000 were calculated with six estimation methods (Table 1A). Method 4 had the lowest coefficient of variation (CV),
0.048, of all the methods. The NOx loads were also calculated for data from 1/90 – 4/00 (Table 1B). The lowest CV for the second set of calculations was 0.052 for Method 6.

Figures 2 and 3 (below) indicate that unusually low NOx concentrations were recorded in 1984. Therefore, load calculations were performed on the data from 1985-2000 (Table 1C). The
lowest CV, 0.043, for Method 4 was lower than any CV obtained in the two NOx load calculations outlined above. Since 1984 represents the oldest data and removing it from the record
improves the CV, the subsequent analyses (with the exception of the section on “concentration versus date”) utilized data from January 1985 to April 2000.

The means and standard deviations of the six methods for load calculations on data from 1/85 – 4/00 are shown in Figure 1.

Concentration versus Date
Figure 2 shows NOx concentration versus date for the entire period. Table 2 provides a statistical summary of the relationship between NOx concentration and date. Figure 3 illustrates
the residuals versus date; the residuals increase slightly with date. Figures 2 and 3 illustrate NOx have been relatively constant with time and that NOx concentrations may have been
unusually low in 1984.

Concentration versus Flow
Figure 4 illustrates NOx concentration versus flow, using Method 4 and data from 1/85-4/00, and Table 3 provides the statistics associated with this graph. Figure 4 demonstrates that
the concentration decreases with flow. Figure 5 illustrates the residuals versus flow; the residuals are constant with flow. Results suggest that the regression approach adequately
describes the flow – concentration relationship.

Date Stratification
The data recorded after 1984 were stratified according to date to determine if there were any detectable trends in NOx concentrations over time. The first stratum included data from
January 1, 1985 – December 1992 while the second stratum included data from January 1993 – April 2000. The load calculations for the six estimation methods are show in Table 4. Method
4 had the lowest CV, 0.040. Figure 6 shows concentration versus flow for this stratification. Figure 6 indicates there was a decrease in NOx concentration between the two periods.
This result is contradictory to the USGS’s finding that nitrate increased in the New Hope Creek near Blands between 1983 and 1995.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The lowest CV, 0.046 was
for Method 4. Figure 7 illustrates concentration versus flow for the seasonal stratification which shows the relationship between concentration and flow is almost the same for the
growing and non-growing seasons.

Flow Stratification
The data were also stratified by flow using 115 cubic hectometers/year as the cutoff value. The lowest CV for this stratification was 0.046 for Methods 2, 3, and 4 (Table 6). Figure
8 illustrates concentration versus flow for the flow stratification using Method 4. Figure 8 illustrates a trendline that decreases steadily for low flows and a trendline that decreases
slowly for high flows. Although this is the same pattern observed for TN flow stratification, the stratification for NOx is not as robust. The graph of concentration versus flow illustrates
a decreasing trend with the exception of four measurements made during high flow events. Although these measurements may indicate that the relationship between NOx concentrations and
flow is different for high flows, there is not enough high flow data to support flow stratification.

Conclusions
Method 4 worked best for estimating NOx loads for all stratification schemes.
Although the date stratification suggested the NOx concentration may have decreased between 1985 and 2000, the USGS found that nitrate concentrations increased in the New Hope Creek
between 1983 and 1995. This determination of an increase may simply be a function of the low concentrations measured in 1984, which were excluded from the bulk of this analysis. There
is not enough evidence to support date stratification or the exclusion of data other than the 1984 measurements.
The relationship between NOx concentration and flow is the same for both the growing and non-growing seasons.
The data suggested the relationship between NOx concentration and flow changes between high and low flows. However, there were not enough NOx measurements at high flows to support
such a stratification.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating NOx loads for New Hope Creek near Blands is:
Use Method 4 for the data from 1/85 – 4/00 with no stratification

Load Calculations
Table 7 shows the annual loads for NOx with Method 4 and no stratification using a maximum interpolation interval of 12 days. Note that FLUX provides both direct ‘model’ outputs and
a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes
in WWTP discharges not accounted for by date stratification. Figure 9 illustrates the time series estimates for this load calculation. Also note that year 2000 results reflect only
the first third of the year, and are therefore not directly comparable to annual totals from previous years.

Table 1A: Calculation of NOx Load on New Hope Creek using Six Methods (84-00)

Table 1B: Calculation of NOx Load on New Hope Creek using Six Methods (90-00)

Table 1C: Calculation of NOx Load on New Hope Creek using Six Methods (85-00) Figure 1: Mean and one Standard Error for Six Estimation Methods
Figure 2: NOx Concentration (µg/l) versus Time (1984-2000)
Table 2: Statistics for NOx Concentration vs. Time
Figure 3: NOx Residuals versus Date (1984-2000)
Figure 4: NOx Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for NOx Concentration vs. Flow
Figure 5: NOx Residuals versus Flow (hm3/yr)

Table 4: NOx Load Calculation with Date Stratification (85-92, 93-00)

Figure 6: NOx Concentration (µg/l) versus Flow (hm3/yr) with Date Stratification
Table 5: NOx Load Calculation with Seasonal Stratification
Figure 7: NOx Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification (Stratum 1 = Oct-Apr, Stratum 2 = May-Sept)
Table 6: NOx Load Calculations for Flow Stratification at 115 hm3/yr
Figure 8: NOx Concentration (µg/l) versus Flow with Flow Stratification at 25 hm3/yr
Table 7: Annual NOx Loads, No Stratification

Figure 9: Time Series of Load Estimates

Total Kjeldahl Nitrogen
FLUX was used to analyze 316 total Kjeldahl nitrogen (TKN) measurements from January 1, 1984 through April 30, 2000. Annual TKN loads for this period were calculated using six estimation
methods (Table 1A). Method 2 had the lowest coefficient of variation (CV), 0.073, of all the methods. Figure 2 (below) illustrates that TKN concentrations may have been unusually
high prior to 1987. The load calculations were repeated using data from January 1, 1987 – April 30, 2000 and the lowest CV for these calculations was 0.046 for Method 2 (Table 1B).
Since the use of the shorter period significantly reduced the CV and eliminated the extreme values reported from 1984 to 1986, this period was utilized in the analysis below unless
otherwise noted.

Concentration versus Date
Figure 2 shows TKN concentration versus date for the entire period (1/84-4/00) using Method 2 and Table 2 provides a statistical summary of the relationship between TKN concentration
and date. Figure 2 illustrates that TKN concentrations were relatively constant over the 16-year period with the exception of the extreme values reported between 1984-1986.

Figure 3 illustrates the residuals versus date for the shorter period between 1/87 – 4/00. The residuals are downward sloping for this period. However, this downward trend may result
from the extreme low values reported in 1999 and 2000.

Concentration versus Flow
Figure 4 illustrates TKN concentration versus flow for data from 1/87-4/00 using Method 2 and Table 3 provides the statistics associated with this graph. Figure 5 illustrates the residuals
versus flow; the residuals decrease slightly with flow, suggesting TKN concentrations decrease slightly as flow increases.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in TKN concentrations with time since 1986. The first stratum contained data from 1/87 –
12/95 while the second stratum contained data from 1/96 – 4/00. Method 4 had the lowest CV for this stratification, 0.040 (Table 4). Figure 6 shows concentration versus flow using
Method 4. Figure 6 shows the trendlines for the two strata crossing in the middle of the data, indicating no significant correlation between TKN concentration and date.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Methods 2, 3 and
4 were the lowest at 0.039. Figure 7 illustrates concentration versus flow for the season stratification. Figure 7 illustrates that TKN concentrations are similar for each season
at low flows, but are noticeably higher for the growing season during high flows. There is enough data to support a seasonal stratification.

Flow Stratification
For the stratification of flows greater than and less than 100 hm3/yr, the lowest CV was 0.049 for Method 2 and the second lowest CV was 0.050 for Method 4 (Table 6). Method 4 was used
to plot concentration versus flow in Figure 9, which illustrates TKN concentrations are relatively constant across low flows and then increase with flow at high flows. This stratification,
however, is not robust because it is heavily reliant on two extreme flow events.

Conclusions
Method 2 was the best method for estimating loads for all stratification schemes except date stratification.
There was no significant correlation between TKN concentration and date (a finding consistent with USGS findings for the period from 1985 to 1995).
The stratification by seasons indicated that TKN concentrations were high during the growing season during high flows. However, the CV for this stratification was the lowest obtained
in the TKN analysis.
Although there may be a change in the relationship between TKN concentrations and flow between low and high flows, the number of measurements during high flows was insufficient to support
flow stratification.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating TKN loads for New Hope Creek near Blands is:
Use Method 2 for the period 1/1/97 – 4/30/00 with seasonal stratification

Load Calculations
Table 7 shows the annual loads for TKN with Method 2 and no stratification using a maximum interpolation interval of 12 days. Note that FLUX provides both direct ‘model’ outputs and
a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes
in WWTP discharges not accounted for by date stratification.

Table 8 shows the annual TKN loads obtained by using a maximum interpolation interval of 12 days and the recommended procedure: Method 2 and seasonal stratification. Figure 10 shows
the time series load estimates obtained by using Method 2 and seasonal stratification. Also note that year 2000 results reflect only the first third of the year, and are therefore
not directly comparable to annual totals from previous years.

Table 1A: Calculation of TKN Load on New Hope Creek using Six Methods (1984-2000)
Table 1B: Calculation of TKN Load on New Hope Creek using Six Methods (1987-2000)

Figure 1: Mean and one Standard Error for Six Estimation Methods (1987-2000)
Figure 2: TKN Concentration (µg/l) versus Date (1984-2000)

Table 2: Statistics for TKN Concentration vs. Date

Figure 3: TKN Residuals versus Date

Figure 4: TKN Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for TKN Concentration vs. Flow

Figure 5: TKN Residuals versus Flow (hm3/yr)
Table 4: TKN Load Calculations for Date Stratification
Figure 6: TKN Concentration (µg/l) versus Flow (hm3/yr) with Date Stratification
(Stratum 1 = 1987-1995; Stratum 2 = 1996-2000)

Table 5: TKN Load Calculations with Seasonal Stratification
(Stratum 1 = Oct-Apr; Stratum 2 = May-Sep)
Figure 7: TKN Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification
Table 6: TKN Load Calculations with Flow Stratification at 100 hm3/yr

Figure 9: TKN Concentration (µg/l) versus Flow with Flow Stratification at 100 hm3/yr

Table 7: Annual TKN Loads from Method 2 with No Stratification
Table 8: Annual TKN Loads from Method 2 with Seasonal Stratification

Figure 10: Time Series of TKN Load Estimates from Method 2 with Seasonal Stratification

Ammonia
FLUX was used to analyze 287 ammonia (NH3) samples from January 1, 1984 through April 30, 2000. The NH3 load for this period was calculated using six estimation methods (Table 1A).
The lowest CV, 0.118, was obtained using Method 4. Figure 2 (below) illustrates unusually high ammonia measurements from 1984 – 1986, so the load calculations were repeated for data
from 1/87 – 4/00 (Table 1B). The lowest CV for these calculations was 0.102 for Methods 2 and 3 while the CV for Method 4 was 0.105. Given that the high measurements record in 1984-1986
were obviously disproportionate to any measurements recorded in the 1990’s and removing 1984-1986 data significantly reduced the CV of all estimation methods, subsequent NH3 analyses
were conducted with data from 1/87 – 4/00, unless otherwise noted. The mean and standard deviation of the six methods calculating loads with data from 1/87 – 4/00 are shown in Figure
1.

Concentration versus Date
Figure 2 shows NH3 concentration versus date for the entire period (1/84 – 4/00) using Method 4 and Table 2 provides a statistical summary of the relationship between NH3 concentration
and date. Figure 2 indicates that usually high ammonia levels were recorded in 1984 – 1986.

Figure 3 illustrates the residuals versus date for the period between 1/87 and 4/00 using Method 2; there is a slight downward slope to the regression residuals during this period suggesting
a decrease in ammonia concentration with time.

Concentration versus Flow
Figure 4 illustrates NH3 concentration versus flow and Table 3 provides the statistics associated with this graph. Figure 4 illustrates the NH3 concentration is not correlated with
flow. Figure 5 illustrates the residuals versus flow. The residuals decrease slightly as flow increases.

Date Stratification
The data recorded after 1986 were stratified according to date to determine if any detectable trends existed in NH3 concentration over time. The first stratum included data from January
1, 1987 – December 31, 1995 while the second stratum included data from January 1, 1996 – April 30, 2000. The load calculations for the six estimation methods are shown in Table 4.
The lowest CV, 0.119, is for Method 3 while the CV for Methods 2 and 4 is 0.124. Figure 6 illustrates concentration versus flow using Method 4 for this date stratification, which
indicates that NH3 concentrations in New Hope Creek have decreased with time between 1987 and 2000. This decrease appears to be more pronounced for concentrations measured during low
flows than it is for concentrations measured during high flows.

Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 3, 0.105,
was the lowest of the estimation methods while the CV for Method 2 was 0.106. Figure 7 illustrates concentration versus flow for the seasonal stratification using Method 2 which shows
that the relationship between concentration and flow is almost the same for growing (May – September) and non-growing seasons (October – April).
Flow Stratification
For the stratification of flows greater than and less than 100 cubic hectometers/year, the lowest CV was 0.101 for Method 2 (Table 6). This stratification was used to plot concentration
versus flow in Figure 8, which illustrates that NH3 decreases with flow. However, there is no change in the relationship between concentration and flow for the two strata; Figure 8
illustrates a simple step decrease between the two trendlines.

Conclusions
Method 2 was the best for estimating NH3 loads for the data from 1/87 – 4/00.
The date stratification suggested a decrease in NH3 concentrations over time between the 1987-1995 period and the 1996-2000 period. However, this decrease was less pronounced for concentrations
at high flows and the USGS found no statistically significant trends in NH3 concentrations over the earlier period of 1983 to 1995.
The seasonal stratification indicated little difference in NH3 concentrations between the growing and non-growing seasons.
Flow stratification is unnecessary because the relationship between NH3 concentration and flow is relatively constant.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating NH3 loads for New Hope Creek near Blands is:
Use Method 2 for the period 1/1/87 – 4/30/00 with no stratification

Load Calculations
Table 7 shows the annual loads for NH3 using Method 2 and no stratification using a maximum interpolation interval of 12 days. Note that FLUX provides both direct ‘model’ outputs and
a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes
in WWTP discharges not accounted for by date stratification. Figure 9 shows the time series estimates for these annual load calculations. Also note that year 2000 results reflect only
the first third of the year, and are therefore not directly comparable to annual totals from previous years.

Table 1A: Calculation of NH3 Loads using Six Estimation Methods (1984-2000)
Table 1B: Calculation of NH3 Loads using Six Estimation Methods (1987-2000)

Figure 1: Mean one Standard Error for Six Methods
Figure 2: NH3 Concentration (µg/l) versus Date

Table 2: Statistics for NH3 Concentration vs. Date
Figure 3: NH3 Residuals versus Date (1987-2000)
Figure 4: NH3 Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for NH3 Concentration vs. Flow

Figure 5: NH3 Residuals versus Flow (hm3/yr)
Table 4: NH3 Load Calculations with Date Stratification
Figure 6: NH3 Concentration (µg/l) versus Flow (hm3/yr) with Date Stratification
(Stratum 1 = 1987-1995; Stratum 2 = 1996-2000)

Table 5: NH3 Load Calculations with Seasonal Stratification

Figure 7: NH3 Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification
(Stratum 1= Oct – April; Stratum 2 = May – September)

Table 6: NH3 Load Calculations with Flow Stratification at 100 hm3/yr
Figure 8: NH3 Concentration (µg/l) versus Flow (hm3/yr) with Flow Stratification
Table 7: Annual NH3 Loads from Method 2 with no Stratification
Figure 9: Time Series of NH3 Load Estimates

Total Phosphorus
FLUX was used to analyze total phosphorus (TP) samples from January 1989 through April 2000. Although data were available from earlier years, previous FLUX analysis confirmed the 1988
phosphorus ban and more stringent point source TP limits established in the 1980s led to a step decrease in phosphorus concentration in the Haw River near Bynum, and a similar trend
would be expected in New Hope Creek. Furthermore, the USGS also reported a step decrease in total phosphorus concentration in the New Hope Creek between 1983 and 1995. The phosphorus
load for this period between 1/89 and 4/00 was calculated using six estimation methods (Table 1A). The lowest CV, 0.053, was obtained using Method 4 (REG-1).

The phosphorus load was also calculated for the data from January 1, 1990 to April 30, 2000 to determine if 1990 was a better start date than 1989 (Table 1B). The lowest CV for the
load calculation was 0.058, again using Method 4. Since the use of the period of 1/89 – 4/00 resulted in a lower CV and includes more data, this period was used in the subsequent analyses
unless otherwise noted.

The mean and standard deviation of the six methods calculating the loads for the period from 1/1/89 – 4/30/00 are shown in Figure 1.

Concentration versus Date
Figure 2 shows TP concentration versus date for the entire period and Table 2 provides a statistical summary of the relationship between total phosphorus concentration and date. Figure
3 illustrates the residuals versus date. Figure 2 illustrates the phosphorus concentration was relatively constant from 1/89 to 4/00, but there were more measurements of low concentrations
from 1997-2000. Figure 3 shows the residuals decreasing with time, suggesting that the phosphorus concentration has decreased slightly with time.

Concentration versus Flow
Figure 4 illustrates phosphorus concentration versus flow using Method 4 and Table 3 provides the statistics associated with this graph. Figure 4 demonstrates that phosphorus concentration
decreases with flow. Figure 5 illustrates the residuals versus flow. The residuals are constant with flow. Results suggest that the regression approach adequately describes the flow-concentration
relationship.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in total phosphorus concentration over time since 1989. The first stratum included data from
January 1, 1989 – December 31, 1995 while the second stratum included data from January 1, 1996 – April 30, 2000. The load calculations for the six estimation methods are shown in
Table 4. Method 4 had the lowest CV, 0.053. Figure 6 illustrates concentration versus flow for this stratification using Method 4. Figure 6 shows phosphorus concentrations decreased
between the two strata. However, the decrease was less pronounced for the high flows.
Seasonal Stratification
The data were stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 6, 0.058,
was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for the seasonal stratification using Method 6. Figure 7 suggests that TP concentrations were
higher in the non-growing season (October – April) for low flows and higher in the growing season for high flows.

Flow Stratification
The cutoff value for the flow stratification that produced the lowest CV and that visually divided the data the best was 100 cubic hectometers/year. The lowest CV for this stratification
was 0.052 for Method 6. Table 6 shows the load calculations for this stratification. A plot of concentration versus flow for this stratification is shown in Figure 8 which indicates
that TP concentrations decrease with flow for both strata, but decrease more gradually with high flows.

Conclusions
Method 4 was best for determining load calculations for all stratification schemes except seasonal and flow stratification.
In the period from 1/89 – 4/00, there was a slight decrease in phosphorus concentrations.
There was no clear relationship between phosphorus concentration and season.
The flow stratification illustrates that the relationship between phosphorus concentration and flow is relatively constant on New Hope Creek.
There was no improvement when the data were stratified by flow because the relationship between flow and concentration (decreasing concentrations with increasing flows) is adequately
explained by the regression model.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating total phosphorus loads for New Hope Creek near Blands is:
Use Method 4 for the period 1/89-4/00 with no stratification

Load Calculations
Table 7 shows the annual loads for total phosphorus from 1/89 to 4/00 with Method 4 and no stratification using a maximum interpolation interval of 12 days. Note that FLUX provides both
direct ‘model’ outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as
might be caused by changes in WWTP discharges not accounted for by date stratification. Figure 9 illustrates the time series load estimates for the load calculations in Table 7. Also
note that year 2000 results reflect only the first third of the year, and are therefore not directly comparable to annual totals from previous years.

Table 1A: Calculation of Total Phosphorus Load using Six Methods (1989-2000)

Table 1B: Calculation of Total Phosphorus Load using Six Methods (1990-2000)
Figure 1: Mean one Standard Error for Six Estimation Methods
Figure 2: TP Concentration (µg/l) versus Date

Table 2: Statistics for TP Concentration vs. Date
Figure 3: TP Residuals versus Date
Figure 4: TP Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for Concentration vs. Flow

Figure 5: TP Residuals versus Flow (hm3/yr)

Table 4: TP Load Calculations with Date Stratification
(Stratum 1 = 1989 – 1995; Stratum 2 = 1996 – 2000)
Figure 6: TP Concentration (µg/l) versus Flow (hm3/yr) with Date Stratification

Table 5: TP Load Calculations with Seasonal Stratification
Figure 7: TP Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification
Table 6: TP Load Calculations with Flow Stratification at 100 hm3/yr

Figure 8: TP Concentration (µg/l) versus Flow with Flow Stratification at 100 hm3/yr

Table 7: Annual Phosphorus Loads with No Stratification

Figure 9: Time Series of TP Load Estimates form Table 10

U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.

5

20

Appendix D/Northeast FLUX.doc

FLUX Analysis of Nutrient Loading
on the Northeast Creek near Genlee
October 1995 – April 2000

May 1, 2001

This document has been prepared by Tetra Tech, Inc. for the
Jordan Lake Nutrient Response Modeling Project.
Funding for this project is provided by:

City of Burlington
City of Graham
City of Greensboro
Town of Mebane
Orange Water and Sewer Authority
Town of Pittsboro
City of Reidsville

Table of Contents

Table of Contents 2
Introduction 3
Total Nitrogen 4
Total Kjeldahl Nitrogen 24
Ammonia 34
Total Phosphorus 44

Introduction

To properly calibrate and verify the Jordan Lake Nutrient Response Model, complete or continuous estimates of tributary nutrient loads must be available or developed. Observed nutrient
data from tributaries consists of point-in-time measurements of concentration, along with continuous estimates of flow from USGS gaging stations. Converting such point-in-time concentrations
to continuous load estimates presents a number of technical challenges because concentration and flow are typically correlated instead of being independent of one another. A variety
of methods are available to address this problem, but no one method is superior for all situations. The US Army Corps of Engineers Waterways Experiment Station produced the FLUX program
to evaluate and compare multiple methods of estimating nutrient loads from tributary concentration data. Tetra Tech used FLUX to analyze nutrient loads for each of the four monitored
tributaries to Jordan Lake. These tributaries are (a) the Haw River; (b) New Hope Creek; (c) Morgan Creek; and (d) Northeast Creek. This report presents the results of the FLUX analysis
for Northeast Creek near Genlee.

The FLUX analysis required input files containing mean daily flows and sample data consisting of measured concentrations of various nutrient species. The daily mean flows of the Northeast
Creek near Genlee were obtained from the USGS (gage 0209741955) and the water quality monitoring data was a combination of USGS data and data from the North Carolina Division of Water
Quality’s ambient water quality monitoring system. The flow gage is located on left bank at downstream side of bridge on Secondary Road 1100, 1.3 miles west of Genlee, and 1.6 miles
downstream of Burdens Creek. Water quality data from both sources were also collected at the SR 1100 bridge crossing.

The data analysis was conducted through April 2000, the most recent month through which monitoring data was available. The start data for the analysis was set to October 1, 1995 because
a gap in USGS flow data for the gage at Genlee exists for the 20 months prior to that date. (Consistent monthly nutrient monitoring data are available as far back as 1971.)

Table 1 below compares the entire flow record from October 1, 1995 – April 30, 2000 to the flow record for the dates on which nutrient measurements were made (sample flows).

Table 1. Flow Data Comparison
Number of Measurements Flow Range
(cfs) Mean Value
(cfs) Median Value
(cfs) Entire Flow Record 1,674 2.4 – 3,350 41 12 Flow on days when nutrient measurements were taken 47 3.5 – 730
54 12
Table 1 indicates that the sample flow values have the same median and a similar average to the entire flow record. However, the sample values do not include values from the high end
of the flow range.

Total Nitrogen

FLUX was used to analyze 47 total nitrogen (TN) samples and 1674 daily flow values from October 1, 1995 through April 30, 2000. The TN load for this period was calculated using six
estimation methods (Table 1). Method 6 (REG-3) was used in the subsequent analysis because it has the lowest coefficient of variation (CV), 0.108, of all the methods. Figure 1 shows
the mean and standard deviations of the six methods.

Concentration versus Date
Figure 2 shows TN concentration versus date for the entire period. Table 2 provides a statistical summary of the relationship between TN concentration and date. Figure 3 illustrates
the residuals versus date; the residuals gradually increase with date. Figures 2 and 3 illustrate that TN concentrations in the Northeast Creek may have increased with time between
1995 and 2000.

Concentration versus Flow
Figure 4 illustrates TN concentration versus flow, using method 6, and Table 3 provides the statistics associated with this graph. Figure 4 demonstrates that TN concentration decreases
with flow. Figure 5 illustrates the residuals versus flow.

Date Stratification
The data was stratified according to date to determine if there was a pronounced difference in TN concentrations in the two halves of the 4.6-year record. The first stratum included
data from October 19, 1995 – December 31, 1997 while the second stratum included data from January 1, 1998 – April 30, 2000. The load calculations for the six estimation methods are
show in Table 4. Method 6 again had the lowest CV, 0.110. Figure 6 shows a plot of concentration versus flow for this stratification. Figure 6 indicates that TN concentrations in
Northeast Creek have increased since 1995.

Seasonal Stratification
The data was stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 6, 0.113,
was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for the season stratification. The graph illustrates a slight increase in TN concentrations
in the non-growing season (October – April), an unexpected result. It should be noted that the monitoring site on Northeast Creek near Genlee is located a short distance downstream
of a major regional wastewater treatment plant (WWTP), the Durham County – Triangle WWTP. Decreased nitrification rates occurring during winter months at the treatment plant, which
result in higher effluent TN concentrations, may account for increased instream TN levels during the non-growing season stratum. However, the limited data, especially during high flows
in the growing season (May – September), make it difficult to draw a conclusion from this analysis.

Flow Stratification
The data were also stratified by flow, testing 10 cubic hectometers/year (hm3/yr) as the cutoff value, based on visual observation of the concentration vs. flow plot (Figure 4). With
flow stratification, the lowest CV was 0.110, resulting from Method 4, while the CV for Method 6 was 0.116 (Table 6). Figure 8 shows a plot of concentration versus flow for stratification
at 10 hm3/yr. Figure 8 illustrates there is no need for flow stratification because the trend for the strata are very similar; the concentrations decrease with flow.

Conclusions
Method 6 was the best for all stratification schemes (and no stratification) except flow stratification.
The CV was lower for no stratification than it was for any of the stratification schemes.
TN concentration appears to have increased gradually with time between 1995 and 2000. However, the limited data makes date stratification unnecessary. (The USGS reported no trends
in the TN concentration at the Northeast Creek near Genlee in the earlier period from 1983 to 1995.)
The seasonal stratification showed an unexpected increase in TN during the non-growing season.
Flow stratification was not needed because the two strata had very similar trends.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating total nitrogen loads for Northeast Creek near Genlee is:
Use Method 6 for the entire period (10/95 – 4/00) with no stratification

Load Calculations
Table 7 shows the annual TN loads for the entire period using the recommended procedure, Method 6 and no stratification. Note that FLUX provides both direct ‘model’ outputs and a modified
estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges
not accounted for by date stratification. Figure 9 illustrates the time series estimates for this load calculation. Also note that results for the years 1995 and 2000 reflect only
partial years, and are therefore not directly comparable to annual totals from full years.

Table 1: Calculation of TN Loads on Northeast Creek using Six Estimation Methods

Figure 1: Mean and one Standard Error for Six Estimation Methods

Figure 2: TN Concentration (µg/l) versus Date

Table 2: Statistics for TN Concentration vs. Date

Figure 3: Plot of TN Residuals versus Date
Figure 4: TN Concentration (µg/l) versus Flow (hm3/yr)

Table 3: Statistics for TN Concentration vs. Flow

Figure 5: TN Residuals versus Flow (hm3/yr)
Table 4: Calculated TN Loads from Six Estimation Methods with Date Stratification
(Stratum 1: 1995 – 1997, Stratum 2: 1998 – 2000)

Figure 6: TN Concentration (µg/l) versus Flow (hm3/yr) with Date Stratification

Table 5: TN Load Calculations with Seasonal Stratification
(Stratum 1: Oct – Apr, Stratum 2: May – Sep)

Figure 7: TN Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification

Table 6: TN Load Calculations with Flow Stratification at 10 hm3/yr

Figure 8: TN Concentration (µg/l) versus Flow with Flow Stratification at 10 hm3/yr

Table 7: Annual TN Load Estimates Oct from Method 6 with No Stratification

Figure 9: Time Series of TN Load Estimates

Nitrate/Nitrite
FLUX was used to calculate nitrate/nitrite (NOx) loads from October 1995 – April using six estimation methods (Table 1). Method 6 was used in the subsequent analysis because it has
the lowest coefficient of variation (CV), 0.143, of all the methods. The mean and standard deviation of the six methods are shown in Figure 1.

Concentration versus Date
Figure 2 shows NOx concentration versus date for the entire period. Table 2 provides a statistical summary of the relationship between NOx concentration and date. Figure 3 illustrates
the residuals versus date; the residuals increase slightly with date. Figures 2 and 3 indicate that NOx concentrations may be increasing with time in Northeast Creek.

Concentration versus Flow
Figure 4 illustrates NOx concentration versus flow, using Method 6, and Table 3 provides the statistics associated with this graph. Figure 4 indicates that concentration decreases with
flow. Figure 5 illustrates the residuals versus flow.

Date Stratification
The data was stratified according to date to determine if there were any detectable trends in NOx concentrations over time. The first stratum included data from October 1995 – December
1997 while the second stratum included data from January 1998 – April 2000. The load calculations for the six estimation methods are show in Table 4. Method 1 had the lowest CV, 0.144
while the CV for Method 6 was 0.157. Since this CV for Method 1 is higher than the CV for Method 6 for no stratification and the data record is short, there is no need for date stratification
and no further analysis was conducted.

Seasonal Stratification
The data was stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The lowest CV, .136 was for
Method 1 while the CV for Method 6 was .143. Figure 6 illustrates concentration versus flow for the season stratification using Method 6 because it is more widely used than Method
1. The graph illustrates a slight increase in NOx concentrations in the non-growing season (October – April), an unexpected result. It should be noted that the monitoring site on Northeast
Creek near Genlee is located a short distance downstream of a major regional wastewater treatment plant (WWTP), the Durham County – Triangle WWTP. The limited data, especially during
high flows in the growing season (May – September), make it difficult to draw a conclusion from this analysis.

Flow Stratification
The data was also stratified by flow using 25 cubic hectometers/year as the cutoff value based on visual observation of the concentration vs. flow plot (Figure 4). With flow stratification,
the lowest CV was 0.131 (Methods 1) while the CV for Method 6 was 0.140 (Table 6). Figure 7 illustrates concentration versus flow for this flow stratification scheme using Method 6.
Figure 7 illustrates there is no need to stratify by flow because the general trend of decreasing concentration with increasing flow is the same for both strata.

Conclusions
Method 6 produced the best results with unstratified data.
The NOx concentrations may indicate a slight increasing trend from 1995 to 2000. However, the USGS reported no statistically significant trend in the NOx concentration in the Northeast
Creek near Genlee in the earlier period from 1983 – 1995.
The seasonal stratification showed an unexpected increase in NOx concentrations during the non-growing season.
Flow stratification was not needed because the two strata had very similar trends.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating NOx loads for Northeast Creek near Genlee is:
Use Method 6 for the entire period (10/95 – 4/00) with no stratification

Load Calculations
Table 7 shows the annual loads for NOx for the entire period with Method 6 and no stratification. Note that FLUX provides both direct ‘model’ outputs and a modified estimator that includes
interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges not accounted
for by date stratification. Figure 8 illustrates the time series estimates for this load calculation. Also note that results for the years 1995 and 2000 reflect only partial years,
and are therefore not directly comparable to annual totals from full years.

Table 1: Calculation of NOx Load on Northeast Creek using Six Estimation Methods

Figure 1: Mean and one Standard Error for Six Estimation Methods

Figure 2: Total NOx Concentration (µg/l) versus Date
Table 2: Statistics for NOx Concentration vs. Date
Figure 3: NOx Residuals versus Date
Figure 4: NOx Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for NOx Concentration vs. Flow
Figure 5: NOx Residuals versus Flow (hm3/yr)

Table 4: NOx Load Calculations with Date Stratification
Table 5: NOx Load Calculations with Seasonal Stratification
(Stratum 1: Oct – Apr, Stratum 2: May – Sep)
Figure 6: NOx Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification

Table 6: NOx Load Calculations with Flow Stratification at 25 hm3/yr
Figure 7: NOx Concentration (µg/l) versus Flow with Stratification at 25 hm3/yr

Table 7: Annual NOx Load Estimates with No Stratification

Figure 8: Time Series of NOx Load Estimates

Total Kjeldahl Nitrogen

FLUX was used to analyze 47 total Kjeldahl nitrogen (TKN) samples from October 1, 1995 through April 30, 2000. The TKN load for this period was calculated using six estimation methods
(Table 1). Method 2 had the lowest coefficient of variation (CV), 0.102, of all the methods and Method 4 had the second lowest CV, 0.105. The means and standard deviation of the six
methods are shown in Figure 1.

Concentration versus Date
Figure 2 shows TKN concentration versus date for the entire period using Method 2. Table 2 provides a statistical summary of the relationship between TKN concentration and date and
Figure 3 illustrates the residuals versus date. Figures 2 and 3 show the TKN concentration in the Northeast Creek was relatively constant over the 4.6-year period.

Concentration versus Flow
Figure 4 illustrates TKN concentration versus flow, using Method 4 and Table 3 provides the statistics associated with this graph. Method 4 was used instead of Method 2 because it better
represented the data visually. Figure 4 indicates the TKN concentration decreases slightly with flow. Figure 5 illustrates the residuals versus flow; the residuals are constant with
flow.

Date Stratification
Since Figures 2 and 3 indicate there is no trend between TKN concentration and date and the data set is small, no attempt was made to stratify the data according to date.

Seasonal Stratification
The data was stratified according to season (October – April and May – September). Table 4 illustrates the load calculations for each estimation method. The CV for Method 2 was 0.091
and the CV for Method 4 was 0.093. Although the CV was lower for Method 2, Method 4 was used in the subsequent analysis because visually it fit the data better. Figure 6 illustrates
concentration versus flow for the season stratification. Figure 6 illustrates that TKN concentrations are slightly greater in the growing season. However, there were only three measurements
corresponding with high flows during the growing season.

Flow Stratification
The data was also stratified by flow using 20 cubic hectometers/year as the cutoff value based on visual observation of the concentration vs. flow plot (Figure 4). With flow stratification,
the lowest CV was 0.098 for Method 6 (Table 5). Method 6 was used to plot concentration versus flow (Figure 7). Although both strata illustrate increasing TKN concentration with increasing
flows, there is a large jump between the two trendlines. This jump signifies that the TKN concentrations are much higher during low flows than during high flows.

Conclusions
There was no clear relationship between TKN concentration and date (a finding consistent with USGS findings for the earlier period from 1985-1995).
TKN concentrations were slightly greater during the growing season. However, since NOx and total nitrogen data both showed high concentrations during the non-growing season and there
was very limited data for the growing season, this is not a robust stratification.
The stratification by flow indicated the TKN concentrations were lower for higher flows.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating TKN loads for Northeast Creek near Genlee is:
Use Method 6 for the period 10/1/95 – 4/30/00 with flow stratification at 20 cubic hectometers/year

Load Calculations
Table 6 shows the annual loads for TKN with Method 4 and no stratification.

Table 7 shows the annual TKN loads obtained by using the recommended procedure: Method 6 and flow stratification at 20. Note that FLUX provides both direct ‘model’ outputs and a modified
estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges
not accounted for by date stratification. Figure 7 shows the time series load estimates obtained by using Method 6 and flow stratification. Also note that results for the years 1995
and 2000 reflect only partial years, and are therefore not directly comparable to annual totals from full years.

Table 1: Calculation of TKN Loads on Northeast Creek using Six Estimation Methods
Figure 1: Mean and one Standard Error for Six Estimation Methods

Figure 2: TKN Concentration (µg/l) versus Date
Table 2: Statistics for TKN Concentration vs. Date
Figure 3: TKN Residuals versus Date

Figure 4: TKN Concentration (µg/l) versus Flow (hm3/yr)

Table 3: Statistics for Concentration vs. Flow

Figure 5: TKN Residuals versus Flow (hm3/yr)
Table 4: TKN Load Calculations with Seasonal Stratification
(Stratum 1: Oct – Apr, Stratum 2: May – Sep)

Figure 6: TKN Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification

Table 5: TKN Load Calculations with Flow Stratification at 20 hm3/yr

Figure 7: TKN Concentration (µg/l) versus Flow with Flow Stratification at 20 hm3/yr

Table 6: Annual TKN Load Estimates from Method 4 with no Stratification

Table 7: Annual TKN Load Estimates from Method 6 with Flow Stratification at 20 hm3/yr

Figure 8: Time Series of TKN Load Estimates from Method 6 with Flow Stratification

Ammonia
FLUX was used to analyze ammonia (NH3) samples from October 1, 1995 through April 30, 2000. The NH3 load for this period was calculated using six estimation methods (Table 1). The
lowest CV, 0.249, was obtained using Method 1. Although this method is not used as much as some of the other methods, it was used in the subsequent analysis because the CV for Method
1 was significantly lower than the CV for any other method. The mean and standard deviation of the six methods are shown in Figure 1.

Concentration versus Date
Figure 2 shows NH3 concentration versus date for the entire period and Table 2 provides a statistical summary of the relationship between NH3 concentration and date. Figure 3 illustrates
the residuals versus date; there is a slight downward slope to the regression residuals during this period.

Concentration versus Flow
Figure 4 illustrates NH3 concentration versus flow and Table 3 provides the statistics associated with this graph. Figure 4 illustrates that NH3 concentration decreases with flow.
Figure 5 illustrates the residuals versus flow. The residuals increase with flow, indicating the downward slope in Figure 4 may be too steep. The inaccurate slope may result from
very limited NH3 measurements during high flow events.

Date Stratification
The data were stratified according to date to determine if there were any detectable trends in NH3 concentration over time. The first stratum included data from October 1, 1995 – December
31, 1997 while the second stratum included data from January 1, 1998 – April 30, 2000. The load calculations for the six estimation methods are shown in Table 4. The lowest CV, 0.252,
was for Method 1. Given that: 1) the CV for the date stratified data was greater than the CV with no stratification; 2) Figures 2 and 3 indicate no clear trend over time, and 3) the
data set is limited, no further analysis was conducted on date stratification schemes.

Seasonal Stratification
The data was stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method with seasonal stratification.
The CV for Method 1, 0.250, was the lowest of the estimation methods. Figure 6 illustrates concentration versus flow for this seasonal stratification scheme using Method 1. The graph
illustrates a slight increase in NH3 concentrations in the non-growing season, an unexpected result. It should be noted that the monitoring site on Northeast Creek near Genlee is located
a short distance downstream of a major regional wastewater treatment plant (WWTP), the Durham County – Triangle WWTP. Decreased nitrification rates occurring during winter months at
the treatment plant, which result in higher effluent NH3 concentrations, may account for increased instream NH3 levels during the non-growing season stratum. However, the limited data,
especially during high flows in the growing season (May – September), make it difficult to draw a conclusion from this analysis.

Flow Stratification
The data was also stratified by flow using 15 cubic hectometers/year as the cutoff value based on visual observation of the concentration vs. flow plot (Figure 4). With flow stratification,
the lowest CV was 0.238 for Method 1 (Table 6). Since the CV for a flow stratification with at cutoff of 15 hm3/yr using Method 1 was the lowest CV obtained in the NH3 analysis, this
scheme was used to plot concentration versus flow (Figure 7). Figure 7 illustrates that NH3 decreases with flow. Since the slopes for the two trendlines are very similar, there is
no need to stratify by flow.

Conclusions
Method 1 was best for determining load calculations with no stratification and all stratification schemes.
There was no clear relationship between NH3 concentration and date between 1995 and 2000. This conclusion is consistent with USGS findings for the earlier period from 1983 to 1995.
The seasonal stratification showed an unexpected increase in NH3 during the non-growing season.
Flow stratification was not needed because the two strata had very similar trends.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating NH3 loads for Northeast Creek near Genlee is:
Use Method 1 for the period 10/1/95 – 4/30/00 with no stratification

Load Calculations
Table 7 shows the annual loads for NH3 using Method 1 and no stratification. Note that FLUX provides both direct ‘model’ outputs and a modified estimator that includes interpolation
on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused by changes in WWTP discharges not accounted for by date stratification.
Figure 8 shows the time series estimates for these annual load calculations. Also note that results for the years 1995 and 2000 reflect only partial years, and are therefore not directly
comparable to annual totals from full years.

Table 1: Calculation of NH3 Loads using Six Estimation Methods

Figure 1: Mean and one Standard Error for Six Methods

Figure 2: NH3 Concentration (µg/l) versus Date

Table 2: Statistics for Concentration vs. Date
Figure 3: NH3 Residuals versus Date
Figure 4: NH3 Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for Concentration vs. Flow

Figure 5: NH3 Residuals versus Flow (hm3/yr)
Table 4: NH3 Load Calculations with Date Stratification
Table 5: NH3 Load Calculations with Seasonal Stratification

Figure 6: NH3 Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification

Table 6: NH3 Load Calculations with Flow Stratification at 15 hm3/yr

Figure 7: NH3 Concentration (µg/l) versus Flow with Flow Stratification at 15 hm3/yr

Table 7: Annual NH3 Load Estimates from Method 1 with No Stratification

Figure 8: Time Series Estimates of NH3 Load Esyimates

Total Phosphorus
FLUX was used to analyze total phosphorus (TP) samples from October 1995 through April 2000. The TP load for this period between was calculated using six estimation methods (Table 1).
The lowest CV, 0.115, was obtained using Method 6. The mean and standard deviation of the six methods are shown in Figure 1.

Concentration versus Date
Figure 2 shows TP concentration versus date for the entire period and Table 2 provides a statistical summary of the relationship between TP concentration and date. Figure 3 illustrates
the residuals versus date. Figures 2 and 3 illustrate TP concentration is relatively constant with date during the period between October 1995 and April 2000.

Concentration versus Flow
Figure 4 illustrates TP concentration versus flow using Method 6 and Table 3 provides the statistics associated with this graph. Figure 4 illustrates that for the data set, TP concentration
decreases with flow. Figure 5 illustrates the residuals versus flow; the residuals are constant with flow.

Date Stratification
The data was stratified according to date to determine if there were any detectable trends in TP concentration over time. The first stratum included data from October 1, 1995 – December
31, 1997 while the second stratum included data from January 1, 1998 – April 30, 2000. The load calculations for the six estimation methods are shown in Table 4. Methods 4 and 6 had
the lowest CV, 0.113. Figure 6 illustrates concentration versus flow for this stratification using Method 6. Figure 6 shows the relationship between concentration and flow is very
similar for the two strata, suggesting no need for date stratification.

Seasonal Stratification
The data was stratified according to season (October – April and May – September). Table 5 illustrates the load calculations for each estimation method. The CV for Method 4, 0.092,
was the lowest of the estimation methods. Figure 7 illustrates concentration versus flow for this seasonal stratification scheme using Method 4. Figure 7 illustrates the TP concentrations
were higher in the non-growing season. However, the limited amount of data collected during the growing season (18 samples during growing season versus 29 during non-growing season)
makes seasonal stratification difficult.

Flow Stratification
Several flow stratification schemes were tested, and the stratification threshold that resulted in the lowest CV and that visually divided the data the best was 25 cubic hectometers/year.
The lowest CV for this stratification was 0.070 for Method 6. Table 6 shows the load calculations for this stratification scheme. A plot of concentration versus flow for this stratification
is shown in Figure 8. Figure 8 shows that the TP concentration is initially constant for flows less than 25 cubic hectometers/year and then decreases as flow increases.
Conclusions
Method 6 was best for determining load calculations for all stratification schemes except seasonal stratification.
In the period from 10/95 – 4/00, there was no clear relationship between TP concentration and date.
Although the data showed the TP concentration was higher during the non-growing season, there is no need for seasonal stratification because the trends for the two seasons were the same
and there is limited data for the growing season.
Flow stratification illustrates that the relationship between TP concentration and flow changes as flows increase on the Northeast Creek. Initially, the concentration is relatively
constant with flow, but then the concentration decreases as flows increase.

Recommended Method and Stratification
Based on the conclusions and findings above the recommended procedure for estimating total phosphorus loads for Northeast Creek near Genlee is:
Use Method 6 for the period 10/95-4/00 with flow stratification at 25 cubic hectometers/year.

Load Calculations
Table 7 shows the annual loads for TP from 10/95 to 4/00 with Method 6 and no stratification.

Table 8 shows the annual TP loads obtained by using the recommended procedure: Method 6 and flow stratification at 25 cubic hectometers/year. Note that FLUX provides both direct ‘model’
outputs and a modified estimator that includes interpolation on residuals. The ‘interpolated’ estimator corrects for un-modeled serial correlation in results, such as might be caused
by changes in WWTP discharges not accounted for by date stratification. Figure 9 illustrates the time series load estimates for the load calculations in Table 8. Also note that results
for the years 1995 and 2000 reflect only partial years, and are therefore not directly comparable to annual totals from full years.

Table 1A: Calculation of Total Phosphorus Load using Six Estimation Methods

Figure 1: Mean and one Standard Error for Six Estimation Methods
Figure 2: TP Concentration (µg/l) versus Date

Table 2: Statistics for Concentration vs. Date
Figure 3: TP Residuals versus Date
Figure 4: TP Concentration (µg/l) versus Flow (hm3/yr)
Table 3: Statistics for Concentration vs. Flow

Figure 5: TP Residuals versus Flow (hm3/yr)

Table 4: TP Load Calculations with Date Stratification
Figure 6: TP Concentration (µg/l) versus Flow (hm3/yr) with Date Stratification

Table 5: TP Load Calculations with Seasonal Stratification

Figure 7: TP Concentration (µg/l) versus Flow (hm3/yr) with Seasonal Stratification

Table 6: TP Load Calculations with Flow Stratification at 25 hm3/yr
Figure 8: TP Concentration (µg/l) versus Flow with Flow Stratification at 25 hm3/yr
Table 7: Annual TP Load Estimates from Method 6 with No Stratification

Table 8: Annual TP Load Estimates from Method 6 with Flow Stratification at 25 hm3/yr

Figure 9: Time Series of TP Load Estimates in Table 8
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.
U.S. Geological Survey. Water-Quality Trends for Streams and Reservoirs in the Research Triangle Area of North Carolina, 1983-95. Water-Resources Investigations Report 97-4061. 1997.

5

2

Jordan Lake Existing Data Memorandum.docContents
Introduction 1
1 Water Quality in Jordan Lake 3
1.1 Monitoring 3
1.2 Physical Parameters 6
1.2.1 Temperature and Thermal Stratification 6
1.2.2 Dissolved Oxygen 7
1.2.3 Solids 8
1.3 Eutrophication Parameters 9
1.3.1 Phosphorus and Nitrogen 10
1.3.2 Chlorophyll a 13
1.3.3 North Carolina Trophic State Index (NCTSI) 16
2 Review of Past Research on Jordan Lake 19
2.1 Findings involving the relationships between nutrients and algal growth 19
2.2 Findings that indicate nutrients alone do not control algal growth 21
2.3 Findings involving the influences of hydrologic events and sedimentation on water quality and algal growth 23
2.4 Findings involving the segmentation and nutrient gradient in Jordan Lake 25
3 Statistical Analysis of Factors Influencing Eutrophication in Jordan Lake 27
4 Algal Community Ecology 33
5 Scoping-Level Model of Lake Nutrient Response 39
5.1 Model Segmentation 40
5.2 Specification of External Loads 41
5.3 BATHTUB Calibration 47
5.4 BATHTUB Model Analysis of Sensitivity of Lake Response to Changing Nutrient Loads 52
6 Initial Review of In-Lake Water Quality Monitoring Program Framework 55
7 Summary and Conclusion 57
8 References 59
Appendix A: Data Summary A-1
Appendix B: Detailed Notes on Literature Review B-1
Appendix C: Principal Components Analysis of Algal Data C-1
Appendix D: FLUX Analyses of Tributary Loads D-1
Appendix E: BATHTUB Model Input Files E-1

List of Figures
Figure 1. Monitoring Stations within Jordan Lake and its Tributaries 5
Figure 2. Vertical Profiles at Station CPF0880A, 1997 6
Figure 3. Dissolved Oxygen Concentrations at Selected Stations in Jordan Lake 7
Figure 4. Correlation of Uncorrected and Corrected Chlorophyll a at CPF086F 11
Figure 5. Total Phosphorus Concentrations, Jordan Lake 12
Figure 6. Total Nitrogen Observations, Jordan Lake 12
Figure 7. Nitrogen-to-Phosphorus Ratios, Jordan Lake 13
Figure 8. Summer Average Uncorrected Chlorophyll a Concentration at Station CPF08C (Lower Morgan Creek Arm). 14
Figure 9. Summer Average Uncorrected Chlorophyll a Concentration at Station CPF0880A (Lower New Hope River Arm). 15
Figure 10. Summer Average Uncorrected Chlorophyll a Concentration at Station CPF055C, Haw River Arm. 15
Figure 11. North Carolina Trophic State Index (NCTSI) for Jordan Lake Stations 17
Figure 12. Model Segmentation 27
Figure 13. Classification and Regression Trees Analysis of Chlorophyll a in Jordan Lake 29
Figure 14. Relationship of Chlorophyll a Concentration to Non-Algal Turbidity 30
Figure 15. Multiple Regression Model of Uncorrected Chlorophyll a in Jordan Lake. 31
Figure 16. Algal Biovolume and NOx Concentrations in New Hope Arm of Jordan Lake (Station CPF087B3), 1983 - 1989. 34
Figure 17. Algal Biovolume and NOx Concentrations in Haw River Arm of Jordan Lake (Station CPF055C), 1983 - 1989. 34
Figure 18. Algal Biovolume Distribution, February 26, 1987 35
Figure 19. Algal Biovolume Distribution, August 25, 1987 36
Figure 20. Algal Biovolume Distribution, August 1, 1989 36
Figure 21. Correlation of Blue-Green Algal Biovolume and Temperature (sample dates: 2/83 – 10/89, 6/96 – 9/97, 4/99 - 8/99) 38
Figure 22. Correlation of Blue-Green Algal Biovolume and NOx Concentration (sample dates: 2/83 – 10/89, 6/96 – 9/97, 4/99 - 8/99) 38
Figure 23. Schematic Representation of BATHTUB Model Network for Jordan Lake 40
Figure 24. Map of Unmonitored Portion of Jordan Lake Watershed 44
Figure 25. Estimated Nitrogen Loading to Jordan Lake, 1997 46
Figure 26. Estimated Phosphorus Loading to Jordan Lake, 1997 46
Figure 27. BATHTUB Nutrient Calibration for Jordan Lake 50
Figure 28. BATHTUB Summer Average Chlorophyll a Calibration for Jordan Lake 51
Figure 29. Sensitivity of Jordan Lake Summer Average Chlorophyll a Concentrations to 25 Percent Changes in Tributary Nitrogen Loads (BATHTUB Analysis Based on 1997 Conditions). 53
Figure 30. Sensitivity of Jordan Lake Summer Average Chlorophyll a Concentrations to 25 Percent Changes in Tributary Phosphorus Loads (BATHTUB Analysis Based on 1997 Conditions). 54
Tables
Table 1. DWQ Tributary Stations 4
Table 2. Dissolved Oxygen Data for Jordan Lake, 1982-1999 8
Table 3. Solids Data for Jordan Lake, 1982-1999 9
Table 4. Average Summer Water Quality, Jordan Lake, 1990-2000 9
Table 5. Nuisance Associations of Dominant Algal Taxa Found in Jordan Lake 37
Table 6. Nonpoint Source Runoff Nutrient Concentrations Assumed for Unmonitored Areas of Jordan Lake Watershed (in mg/l) 43
Table 7. Estimated Nutrient Loads for Jordan Lake BATHTUB Analysis, 1997 45
Table 8. Calibration Factors for Jordan Lake BATHTUB Model 48
Table 9. Summary of Nutrient Response Related Water Quality Monitoring Efforts in Jordan Lake, Showing Monitoring Agency, Parametric Coverage, and Approximate Sampling Frequency 56

Introduction
Seven local governments with wastewater treatment facilities in the Jordan Lake Watershed (the “Project Partners”) have agreed to work together with the State and watershed stakeholder
organizations to develop a nutrient response model for B. Everett Jordan Reservoir (“Jordan Lake”). The lake is considered by the NC Division of Water Quality to be one of the most
eutrophic lakes in the State, and legislation enacted in 1997 requires wastewater treatment facilities in the lake’s watershed to meet a total nitrogen limit of 5.5 mg/l by January
1, 2003. This is a generic limit applicable to all designated Nutrient Sensitive Waters, and is not based on an analysis of nutrient response in Jordan Lake. NC General Statute 143-215.1
provides for, as an alternative approach, the determination of site-specific wasteload allocations through use of a calibrated in-lake nutrient response model. The nutrient response
model must be capable of predicting the impact of nitrogen and phosphorus loading to lake water quality, and will be used to evaluate the potential for alternative nitrogen limits for
the wastewater treatment facilities. Accordingly, the Project Partners retained water quality modeling services of Tetra Tech, Inc. to address this need. The model must be developed,
calibrated, verified, and delivered to the Project Partners by December 1, 2001, with submission of results to the NC Environmental Management Commission by April 2002.
This technical memorandum is the first step in the creation of the nutrient response model. It provides a summary of the existing data on water quality and nutrient response within
Jordan Lake with a focus on temperature, dissolved oxygen, solids, nutrients, and chlorophyll a. Past water quality monitoring has included the analysis of metals, organics, and other
parameters; however, those data are not reviewed in this report as they do not directly relate to nutrient loading and response. Sections 1 through 6 of the memorandum document the
findings of the project team’s initial analyses as follows:
Water Quality in Jordan Lake: A summary of the existing monitoring programs and associated data within the lake.
Review of Past Research on Jordan Lake: A summary of the key results of past investigations of the lake, focusing on nutrients and algal response.
Statistical Analysis of Factors Influencing Eutrophication in Jordan Lake: An examination of the relationship between observed chlorophyll a concentrations and potential causal factors
from an empirical, statistical viewpoint.
Algal Community Ecology: An initial examination of the relationship between algal response and potential causal factors from an ecological, process-based perspective.
Scoping-Level Model of Lake Nutrient Response: An exploratory evaluation of the processes controlling algal response using a simple scoping model.
Initial Review of In-Lake Water Quality Monitoring Program Framework.
Finally, Section 7 of the memorandum summarizes how the results of these analyses lay the foundation for selection of a final modeling tool to create the calibrated in-lake nutrient
response model required by NC General Statute 143-215.1.

Water Quality in Jordan Lake
An extensive amount of water quality data has been collected for Jordan Lake. A list of files and the data they contain are summarized in Appendix A. This section of the memorandum
describes principal monitoring programs, and summarizes key findings for water quality parameters of primary interest to the lake modeling project.
Monitoring
Since B. Everett Jordan Reservoir was impounded in 1982 it has been monitored extensively by a number of state and federal agencies, as well as local governments that use water from
the reservoir, or discharge wastewater in its immediate vicinity. The North Carolina Division of Water Quality (formerly the Division of Environmental Management) has performed the
most extensive collection of data. The U.S. Army Corps of Engineers also collected substantial water quality data on the Reservoir during its first 5 years in existence. In more recent
years the U.S. Geological Survey, through the Triangle Area Water Supply Monitoring Project, has regularly monitored water quality in the reservoir. Local governments that have performed
water quality monitoring include the Town of Cary, the City of Durham, Orange Water and Sewer Authority (OWASA), the Town of Pittsboro, and Durham County.
Primary water quality monitoring by the N.C. Division of Water Quality (DWQ) has generally consisted of depth stratified data for physical parameters, photic zone data for nutrients
and chlorophyll a, occasional bottom data for nutrients, and surface data for metals. During the first five years after flooding of the lake, from July 1982 to October 1987, DWQ performed
such water quality monitoring monthly, year round, at numerous monitoring stations (refer to map in Figure 1). The sampling schedule was reduced to monthly collection during growing
season months (April – October) from 1988 through 1990. DWQ monitoring frequency was reduced further to 1-2 sampling events annually from 1991-1995, and monthly growing season (May-
August) sampling was resumed in 1996 and carried through 1999. In conjunction with this study, year 2000 sampling frequency has been increased to twice per month from May to September.
DWQ has also performed extensive algal monitoring in the reservoir. Collection of algal data has generally occurred along the same schedule and with the same frequencies as those described
above for the collection of physical and chemical data with the exception of the 1990-1995 time frame, during which no algal monitoring was performed. Algal community data available
from DWQ for Jordan Lake include counts and biovolumes by individual species for all years in which data were collected. DWQ has also performed algal growth potential tests (AGPT)
on samples from Jordan Lake in 1985 at one location, 1989 at two locations and in 1996 at five locations.
The Town of Cary has also collected algal data since 1998 at their surface water intake site, located just north of U.S. Highway 64 on the east bank of the lake (refer to map in Figure
1). However, Cary’s data is limited to three particular types of algae that can adversely affect drinking water quality, Anabaena sp., Synedra sp., and Melosira sp.
In addition to in-lake data, DWQ has collected monthly water quality data for all of the major tributaries to Jordan Lake from monitoring sites immediately upstream from the lake (Table
1). These are shown in Figure 1 as open circles with crosshairs. The tributary data were collected through DWQ’s ongoing statewide ambient water quality monitoring program in which
monthly monitoring is performed at run-of-river monitoring sites across North Carolina for a wide range of physical, chemical and biological parameters. Data from Jordan Lake tributary
sites for 1976 through 1999 are available for download from the U.S. Environmental Protection Agency’s data storage and retrieval system (STORET).
DWQ Tributary Stations
Station Name USGS DWQ STORET Morgan Creek near Farrington 02097521 CPF086 B3900000 New Hope Creek near Blands 02097314 CPF074J B3040000 Northeast Creek near Genlee 0209741955 CPF081 B3660000 Robe
rson Creek at SR 1939 02097189 CPF055 B2450000 Haw River near Bynum 02096960 CPF049 B2100000
In recognition of the fact that the communities of the Triangle Region of North Carolina are heavily dependent on surface water for drinking water supplies, several local governments
from the region joined to form the Triangle Area Water Supply Monitoring Project (TAWSMP). In partnership with and on behalf of the TAWSP, the U.S. Geological Survey (USGS) began collecting
water quality data at numerous sites throughout the region in October 1988. Thirteen of the TAWSMP sites are located in Jordan Lake proper or its immediate tributaries (seven are within
the extent on the map in Figure 1). Data from the sites includes the wide range of physical and chemical parameters typically monitored by USGS.
Three major municipal wastewater treatment plants (South Durham Water Reclamation Facility, Durham County Triangle Wastewater Plant, and OWASA Mason Farm Wastewater Plant) discharge
to Jordan Lake tributaries at locations immediately upstream of the New Hope arm of the lake (refer to map in Figure 1). The NPDES discharge permits for these facilities stipulate that
they monitor water quality at sites in and above the reservoir for physical parameters and nutrients. In-stream monitoring data for Roberson Creek, which discharges to the Haw River
arm of the lake, is also available from the Town of Pittsboro Wastewater Treatment Plant, as the Town is required to conduct limited monitoring both upstream and downstream of that
facility.

Monitoring Stations within Jordan Lake and its Tributaries
Physical Parameters
Temperature and Thermal Stratification
Jordan Lake experiences annual cycles of water temperature that track average atmospheric temperature, with summer surface water temperatures generally in the range of 28 to 31( Celsius.
As in most Piedmont lakes, summer conditions are typically accompanied by thermal stratification, in which warm surface water overlies colder, denser bottom water. When thermal stratification
is present it impedes mixing and limits reaeration of the bottom water. During stratification, oxygen is depleted in the bottom water. Under those conditions, sediment-bound nutrients
and metals may be released, thereby leading to elevated concentrations of nutrients and metal ions.
Some typical vertical profiles from Station CPF0880A (lower New Hope arm) during the summer of 1997 are shown in Figure 2. On June 9, 1997 the water column was unstratified, with nearly
isothermic temperatures across all depths. The dissolved oxygen (DO) profile also shows evidence of mixing, as DO declines only very near the bottom, where the influence of sediment
oxygen demand is present. By July 1 of that year, stratification had developed, as shown by the relatively well-defined thermocline in the temperature profile in the region of 5 m
depth. Temperatures below this point are near 20( C, or little changed since June. Below the thermocline, almost all of the dissolved oxygen has been depleted from the water column
as mixing from the surface layer is impeded. The profile of August 4, 1997 shows results of partial mixing down to about 8 m depth. Stratification further declined in September.

Vertical Profiles at Station CPF0880A, 1997
In general, stratification is not particularly robust in Jordan Lake. While there is plenty of summer warming of the surface water to promote stratification, other factors appear to
promote mixing. Three of the factors promoting mixing are hydraulic mixing due to storm inflows, turbulent mixing caused by heavy boat traffic, and the alignment of the New Hope arm
with the prevailing summer southwesterly winds that promote wind mixing. Thermal stratification is often present in mid-summer, but may form and break down several times during a season.
Within the shallow upper portion of the New Hope arm, thermal stratification is rarely present in DWQ observations.
Dissolved Oxygen
Figure 3 summarizes observations of dissolved oxygen in surface water at four stations in Jordan Lake. Dissolved oxygen concentrations have generally remained above the state water
quality standard of 5 mg/L, with a few exceptions. Most notably, on 24 Sept. 1996 the dissolved oxygen concentration was 1.7 mg/L at CPF086C, 5.6 mg/L at CPF0880A, 4.7 mg/L at CPF055C
and 4.1 mg/L at CPF055E. These observations were taken in the aftermath of Hurricane Fran, and represent the combined effects of the washoff of oxygen demanding materials into the
lake and mixing of oxygen-depleted bottom waters by the storm. A few other observations less than 5 mg/L were recorded prior to 1990. It should be noted, however, that the observed
data are primarily later morning observations that do not reflect the minimum of the daily dissolved oxygen cycle caused by algal respiration overnight.
Dissolved Oxygen Concentrations at Selected Stations in Jordan Lake

Due to intense algal production in the lake, it is more common to observe excess dissolved gasses in the surface water than depleted dissolved oxygen. North Carolina water quality regulations
(NCAC 2B.0211 (b)(E)) specify total dissolved gasses “not greater than 110 percent of saturation.” Due to algal production of dissolved oxygen, this criterion is often exceeded during
the summer in Jordan Lake. Table 2 summarizes surface DO concentration and percent saturation data for the period of record.
Dissolved Oxygen Data for Jordan Lake, 1982-1999
Parameter Station Average Median Maximum Minimum Dissolved Oxygen (mg/L) CPF086C 9.4 9.1 13.6 1.7 CPF0880A 8.9 8.8 12.8 5.5 CPF055C 10.3 10.4 16.8 4.7 CPF055E 9.7 10.0 17.5 4.1 Percent
Saturation CPF086C 103.7 103.2 153.2 19.6 CPF0880A 98.3 98.3 140.5 61.3 CPF055C 118.4 109.6 214.4 55.3 CPF055E 108.0 101.2 221.9 48.2
In contrast to the surface, dissolved oxygen is often depleted in bottom waters of the lake when stratification is present (see Figure 2-2 above). With mixing impeded by thermal stratification,
bacterial consumption of organic material in the water column and bottom sediment quickly depletes the dissolved oxygen. This phenomenon of hypolimnetic oxygen depletion occurs most
frequently at the deeper stations in the lower New Hope Arm and near the dam, where stratification is more persistent. 32 percent of the observations below 5 m depth at CPF0880A and
34 percent of the observations below 5 m depth at CPF055E had less than 1 mg/L dissolved oxygen. In contrast, the percentage of time during which DO was less than 1 mg/L near the bottom
at CPF086C in the upper New Hope Arm was only 9 percent.
Solids
Solids, both suspended and dissolved, can play an important role in determining lake response to nutrient loads because solids limit light penetration into the water column, and thus
also limit algal growth. Settling solids may also remove particle-reactive materials, such as orthophosphate, from the water column. Solids concentrations in Jordan Lake are summarized
below in Table 3. Solids are relatively high, and include a large component of dissolved or colloidal solids (dissolved solids are usually not monitored directly, but may be determined
from the difference between total solids and total suspended solids). Concentrations are highest in the upper New Hope Arm segment.

Solids Data for Jordan Lake, 1982-1999
Parameter Station Average Median Maximum Minimum Total Suspended Solids (mg/L) CPF086C 25.6 21 94 4 CPF0880A 8.1 7 63 1 CPF055C 12.0 10 60 1 CPF055E 9.4 8 38 2 Total Solids (mg/L) CPF086C 216
.7 130 7600 65 CPF0880A 133.0 100 2900 50 CPF055C 140.2 140 230 88 CPF055E 126.0 120 200 85
Eutrophication Parameters
Eutrophication is a scientific term for the condition of nutrient over-enrichment of a waterbody that leads to excessive growth of algae and other aquatic plants. Nutrients such as
phosphorus and nitrogen are essential for the growth of plant life, including aquatic plants and algae. However, at high concentrations these nutrients can over-enrich waterbodies
and result in degraded water quality.
Algal blooms occur at various times throughout the year; however, water column concentrations of the nutrients nitrogen and phosphorus are of greatest concern for causing excess algal
growth and undesirable eutrophication during the summer growing season. Table 4 summarizes summer average water quality conditions in recent monitoring (1990—2000).
Average Summer Water Quality, Jordan Lake, 1990-2000
Station Number of Samples pH Conductivity (µS/cm) Secchi Depth (m) Total P (mg/L) Total N (mg/L) N:P Ratio Uncorrected Chlorophyll a
(µg/L) CPF081A1C 27 8.3 144 0.40 0.093 0.61 7.8 49 CPF086C 27 8.1 144 0.42 0.080 0.60 8.0 44 CPF086F 27 8.1 138 0.46 0.068 0.59 9.1 48 CPF087B3 25 7.7 129 0.73 0.032 0.48 19.8 22.5 CPF0880A 27 8
.0 131 0.86 0.029 0.44 18.5 21 CPF055C 27 8.9 198 0.55 0.087 0.67 8.8 36 CPF055E 29 8.7 165 0.70 0.050 0.56 12.6 28
In the table, stations CPF081A1C through CPF0880A are arranged from upstream to downstream in the New Hope arm. CPF055C is in the Haw River arm, while CP055E is just above the dam.
On average, concentrations of nitrogen and phosphorus decrease with distance down the New Hope arm (and away from point sources), as do average chlorophyll a concentrations, while
Secchi depth (water clarity) and pH increase. Nutrient concentrations in the Haw River arm are substantially higher, on average, than those in the adjacent lower part of the New Hope
arm.
Note that the chlorophyll a values reported in the table are “uncorrected” for the presence of pheophytin a, a degradation product of chlorophyll a. Corrected values are generally preferable
as indicators of live algal concentration, and North Carolina’s chlorophyll a water quality standard is defined in terms of the corrected concentration. Unfortunately, there are analytical
and quality control problems with much of the NC DWQ historical database for chlorophyll a that cause these data to be unusable for certain time periods (personal communication from
Jay Sauber, NC DWQ, 2001). USGS has also collected chlorophyll a data at various locations in Jordan Lake, but the method of analysis is different from DWQ and would not be an accurate
comparison in a trend analysis. Therefore, trends over time are best examined using only the uncorrected data from DWQ.
The correction for pheophytin a is used because it absorbs and fluoresces light in the same range of the spectrum as chlorophyll a (APHA, 1985). For fluorometric determination of chlorophyll
a concentration, measurements are taken before and after acidification of the sample. The ratio of these two measurements (the acid ratio) is then used to calculate corrected chlorophyll
a concentrations (without interference from pheophytin a). Unfortunately, the acid ratio can only be calculated at the time of lab analysis: “the acid ratio is instrument specific
and must be determined for each instrument at the time of calibration” (Baker et al., 1983). This ratio was incorrectly calculated after September 1996 through 2000, so corrected
values are not recoverable for this period.
In addition to problems with the acid ratio, the calibration standard, used in the calculation of both uncorrected and corrected values, was also calculated in error for September 1996
through 2000. This mistake is easily remedied for uncorrected values by recalculation with the appropriate standard. These uncorrected chlorophyll a values have been recalculated
by DWQ and are used with values from earlier years to represent trends in Jordan Lake.
Prior to September 1996, the uncorrected and corrected chlorophyll a data generally track well for most of the stations at Jordan Lake, indicating that the correction for pheophytin
a is generally small. Figure 4 shows these data for station CPF086F, which is located in the downstream portion of the Upper New Hope segment (Figure 1). The correlation coefficient
for these two data sets is 0.81. Given the high correlation, it will be appropriate to use the uncorrected data for model calibration and trend analysis.
Phosphorus and Nitrogen
Jordan Lake receives discharges from a number of wastewater treatment plants, including three major dischargers to the upper New Hope arm and seven major dischargers to the Haw River.
There are also many minor discharging facilities (less than 1 MGD permitted flow), including the Town of Pittsboro’s wastewater plant, and a number of privately owned discharge facilities.
The lake also receives significant runoff from urban and suburban developed areas. As a result, the nutrient loading to the lake is relatively high. During the early years after
impoundment, nutrient concentrations were further enhanced due to lake start-up effects caused by the decay of residual organic material in the lakebed.

Correlation of Uncorrected and Corrected Chlorophyll a at CPF086F

Figure 5 and Figure 6 provide the complete time series of observed total phosphorus and total nitrogen concentrations at four selected monitoring stations in the lake. Stations CPF086C
and CPF0880A represent the upper and lower New Hope arm, respectively. A general declining trend in both phosphorus and nitrogen is evident at both stations, with concentrations after
1990 lower than those prior to 1990. In part, this may be an artifact of the change from year-round monthly sampling in 1982-1987 to summer-only sampling in subsequent years. Station
CPF055C is in the Haw River arm, and shows concentrations higher than those seen in the New Hope arm. Concentrations near the dam (CPF055E) closely follow those seen in the Haw River
arm.
It is also of interest to examine the summer average nitrogen to phosphorus (N:P) ratios. Algae require these nutrients in specific ratios to support cell growth, so the N:P ratio is
often used as an indicator of the nutrient that is more limiting to algal growth—i.e., the nutrient for which response to increased loads would be the most significant. Optimal N:P
ratios vary somewhat by algal taxa, but a ratio less than 10 is generally considered to indicate nitrogen limitation, while a ratio greater than 15 suggests phosphorus limitation.
Within the New Hope arm, the average summer N:P ratio also increases downstream. In the upper segment, above SR 1008, the N:P ratio suggests nitrogen limitation. Downstream of SR
1008, however, the ratio suggests phosphorus limitation. This reflects proportionately greater loss rates of phosphorus than nitrogen with transport downstream. These average results
do not tell the whole story, however, as N:P ratios vary considerably among sampling dates, and suggest that sometimes nitrogen and sometimes phosphorus may be limiting. This is shown
in Figure 7, which displays the time series of N:P ratios.

Total Phosphorus Concentrations, Jordan Lake

Total Nitrogen Observations, Jordan Lake
Nitrogen-to-Phosphorus Ratios, Jordan Lake

Chlorophyll a
Chlorophyll a is the primary photosynthetic pigment in most algae and is often used as an indicator of algal density. As noted previously, chlorophyll a concentrations in Jordan Lake
exhibit a spatial gradient along the New Hope arm, with highest concentrations in the upstream source areas, and lower concentrations downstream. There are also temporal patterns in
the data. First, there is a seasonal component, with highest concentrations usually found in the warm summer period (June through September). In addition, concentrations in the year
immediately after impoundment were higher than those observed later in the New Hope arm. For example, Figure 8 shows summer average results by year at CPF086C (lower Morgan Creek arm),
where the 1982 average chlorophyll a concentrations approached 190 µg/L. Such start-up effects are common for new reservoirs due to the decay of terrestrial organic material (with
associated nutrient release) following filling of the reservoir.
Again note that uncorrected chlorophyll a data are used to represent trends in Jordan Lake due to problems with corrected values. Stations CPF086C and CPF081A1C have no reported values
for 1988.

Summer Average Uncorrected Chlorophyll a Concentration
at Station CPF086C (Lower Morgan Creek Arm).

Trends are less evident in the later data. However, concentrations prior to 1990 appear to have been on average higher than those seen after 1990. This is likely caused by the net
impacts of a number of factors, including diminishing influence of lake start up effects, the phosphate detergent ban in 1989, and improvements in OWASA’s Mason Farm wastewater treatment
plant on Morgan Creek (upgrade to 8.0 MGD with biological nutrient removal completed in Fall 1991).
Figure 9 shows summer average chlorophyll a concentrations further down the New Hope arm at CPF0880A. The pattern here is similar to that seen at CPF086C—emphasizing the importance
of upstream sources including South Durham Water Reclamation Facility, Durham County Triangle Wastewater Plant, and OWASA Mason Farm Wastewater Plant —although the concentrations at
this station are lower. Conditions are also affected by the phosphate detergent ban and upgrades to biological nutrient removal at OWASA Mason Farm WWTP in 1991 and at South Durham
(New Hope Creek) in 1994.
The temporal patterns in the Haw River arm (CPF055C; Figure 10) differ somewhat from those in the New Hope arm, reflecting the shorter residence time and more river-like character of
this part of the reservoir. There is little or no start up effect evident at this station. As with the New Hope stations, there does appear to have been a small decrease in chlorophyll
a concentration after about 1989.

Summer Average Uncorrected Chlorophyll a Concentration
at Station CPF0880A (Lower New Hope River Arm).

Summer Average Uncorrected Chlorophyll a Concentration
at Station CPF055C, Haw River Arm.
North Carolina Trophic State Index (NCTSI)
The State of North Carolina developed a numerical index of lake eutrophication status, known as the North Carolina Tropic State Index or NCTSI, as part of the State’s original Clean
Lakes Classification Survey (NRCD, 1982). This index has been used to compare and track status of North Carolina lakes, and takes into account observations of total organic nitrogen
(TON, mg/L), total phosphorus (TP), Secchi Depth (SD, inches), and corrected chlorophyll a. The index is calculated as:
NCTSI = TON Score + TP Score + SD Score + CHL Score
where
TON Score = [Log(TON) + 0.45]/0.24 x 0.90
TP Score = [Log(TP) + 1.55]/0.35 x 0.92
SD Score = [Log(SD) – 1.73]/0.35 x –0.82
CHL Score = [Log(CHL) – 1.00]/0.43 x 0.83

Lower NCTSI scores represent better water quality conditions. The NCTSI is interpreted to trophic state as follows:
< -2.0 Oligotrophic
-2.0 to 0.0 Mesotrophic
0.0 to 5.0 Eutrophic
> 5.0 Hypereutrophic
NCTSI scores for six Jordan Lake stations are shown in Figure 11, based on the average of June through September observations in each year. Spatially, the scores are consistently higher
in the Upper New Hope and Haw River Arms than in the Middle and Lower New Hope, indicating more eutrophic conditions. At all stations, the scores appear to have decreased (improved)
somewhat up to the early 1990’s, but have remained relatively constant since that time. NCDEM (1992) also prepared lake-wide average scores for Jordan Lake through 1990. These range
from a low of 3.3 in 1984 to a high of 5.7 in 1986. A summary of “most recent” NCTSI scores through 1990 (NCDEM, 1992) shows Jordan Lake with the 11th highest score out of 138 lakes
ranked. Of lakes in the Piedmont Region, only Lake Lee (Monroe), Lake Fisher (Concord), and Lake Crabtree (Raleigh) had higher NCTSI scores than Jordan Lake.
Current NCTSI scores for the Upper New Hope and Haw River arm are mostly in the 3 to 4 range, indicating strongly eutrophic conditions. Scores near the dam range from 2 to 3, while
those in the Middle New Hope range from 0 to 2, indicating mildly eutrophic conditions.
Scores for 1999 and 2000 have been omitted due to problems with the corrected chlorophyll a data. Corrected chlorophyll a data from 1996 to 1998 are probably usable, although likely
subject to random "noise" due to poor quality control measures (Sauber, 2001).

North Carolina Trophic State Index (NCTSI) for Jordan Lake Stations

Review of Past Research on Jordan Lake
Detailed research on water quality and eutrophication in Jordan Lake has been conducted by a number of authors. A review of the available literature identified the key reports for evaluation
of algal response as those of Weiss et al. (1984, 1985), Kuenzler et al. (1986) and Jackson et al. (1992). Each of these reports contains valuable insights into the factors controlling
algal growth in Jordan Lake. It should be noted, however, that this research focuses on the earlier post-impoundment period, during which lake start-up effects are likely to have been
dominant. Therefore, the conclusions provided by these authors need to be verified in more recent data.
Key findings from past research are summarized below in bullet form. A more detailed review of past research is provided in Appendix B.
Findings involving the relationships between nutrients and algal growth
There is a negative relationship between nitrogen and phosphorus and algal growth.
Nitrogen and phosphorus concentrations are highest in winter and “lowest in early and late summer when assimilation by phytoplankton peaks” (Jackson et al. 1992, p. 11).
“Correlations with nutrient concentrations were negative, indicating that algal growth was controlling nutrient concentrations rather than nutrients controlling algae” (Kuenzler et al.
1986, p. xvi).
While chlorophyll a was positively correlated with particulate nitrogen (PN), both chlorophyll a and PN were negatively correlated with filterable reactive phosphorus (FRP) and chlorophyll
was negatively correlated with nitrate. The negative correlations indicate the “effective removal of nutrients during times of phytoplankton abundance” (Kuenzler et al. 1986, p. 77-79).
Neither phosphorus nor nitrogen consistently limited algal growth in Jordan Lake.
Dissolved inorganic nitrogen to dissolved inorganic phosphorus (DIN:DIP) ratios “suggest that algal growth is frequently limited by P” (Kuenzler et al. 1986, p. 100).
“Algal assays conducted on water from Jordan Lake showed that neither N nor P alone were consistently at concentrations low enough to limit further phytoplankton growth” … “Therefore
it appears that reduction of either nutrient … will contribute to control of algal growth” (Kuenzler et al. 1986, p. 101-103).
“The TN:TP values suggested mixed limitation” (Weiss et al. 1985, p. 52).
DWQ conducted Algal Growth Potential Tests (AGPT) during 1999 in the Haw River and lower New Hope arms of Jordan Lake. The results indicated that algal growth at all five stations was
limited by nitrogen availability (NC DWQ 1999, p. 20).
Nutrient cycling is very important to algal growth.
“Nutrient dynamics in lake waters cannot be understood simply by measuring concentrations. Periods of maximum biotic activity often correspond to periods of low P concentrations that
should, if recycling is ignored, limit further biological activity.” This indicates the importance of measuring nutrient flux. (Kuenzler et al. 1986, p. 6).
“Ambient dissolved inorganic N concentrations can be very low when a dynamic balance exists between ammonium removal and regeneration” (Kuenzler et al. 1986, p. 12).
“OP [orthophosphate concentration] was quite low or below detection limits during much of the growing season in the New Hope arm suggesting phosphorus recycling” (Weiss et al. 1985,
p. 52).
The low concentrations of biologically available forms of N and P suggest that recycling of one or both nutrients is necessary to maintain phytoplankton growth (Weiss et al. 1986, p.
43).
Ammonium is the nitrogen species preferred by algae.
“Ammonia is made available as a nutrient to the phytoplankton by regeneration from organic matter, by zooplankton excretion, bacterial remineralization, or leakage from photoplankton
cells themselves (Brezonik 1972). Primary production associated with ammonium assimilation is therefore termed ‘regenerated production’” (Kuenzler et al. 1986, p. 12)
Ammonia is favored as N source for phytoplankton since it does not require reduction as nitrate does (Kuenzler et al. 1986, p. 12).
The following studies on regenerated N on lake systems “indicate that ammonia assimilation is very important and may even control primary production, especially during algal blooms”:
Alexander (1970), Brezonik (1972), Toetz and Cole (1980), Axler et al. (1981) (Kuenzler et al. 1986, p. 12).
“Ammonium uptake rates in the Haw River arm steadily increased from May until October when rates peaked at 26 g / L-hr.” The New Hope arm also had high rates during this period. “Nitrate
uptake was also rapid, but generally slower than ammonium uptake.” (Kuenzler et al. 1986, p. 61).
A log-log graph of the Relative Preference Indices “shows algal preference is predominantly for NH4 over NO3 at all stations.” (Kuenzler et al. 1986, p. 98).

Bacterial uptake of phosphorus may also be an important consideration.
Kuenzler and Greer (1980) “showed that phytoplankton is not the only agent removing phosphorus from the water; for most of the year ‘bacterial’ uptake and uptake by suspended sediments
was very important.” (Kuenzler et al. 1986, p. 3).
“Biotic uptake [of phosphorus] was dominated by the small size fraction [0.45 to 8.0 m fraction, initially assumed to be mostly bacterial], averaging 87.6 ( 7.9% of total biotic uptake”
(Kuenzler et al. 1986, p. 41).
“About 90% of the biotic uptake was insensitive to antibiotics in March and 58% in May, suggesting that most [phosphorus] uptake was procaryotic” (Kuenzler et al. 1986, p. 48).
“Small particle uptake was well correlated with large particle uptake at Station 5 [upper New Hope arm] (r=0.88) and Station 10 [middle New Hope arm] (r=0.73), suggesting that algal
and bacterial uptake increased or decreased together and that algae did not dominate uptake during periods of high biomass and P-deficiency.” (Kuenzler et al. 1986, p. 83).
Algae can utilize internal phosphorus sources when the concentration is low in waters.
“Many phytoplankton species can take up P in excess of their metabolic needs and store it within the cell as polyphosphate (Perry 1976)” Algae can use this internal source when P is
limited in water (Kuenzler et al. 1986, p. 5-6).
Findings that indicate nutrients alone do not control algal growth
Efforts to model algal growth using only nutrients as the controlling factors have been relatively unsuccessful.
“Weiss, et al. (1985) employed the Dillon-Rigler model, which relates total phosphorus inputs to phytoplankton production, as measured by chlorophyll a, and found predictions to consistently
overestimate algal production in the New Hope arm.” They concluded total phosphorus was only one of several possible factors contributing to algal growth (Jackson et al. 1992, p. 20).
“Our attempt to use the Dillon-Rigler model to predict chlorophyll a for B. Everett Jordan Lake using data from Years I and II, identified a variation of five orders of magnitude for
the same TP concentration. This suggests that TP at least is not the only factor controlling chlorophyll a.” (Weiss et al. 1985, p. xviii).
Inorganic suspended solids, temperature, turbidity, and other physical parameters influence algal growth.
“Kuenzler et al. (1986) found that neither phosphorus nor nitrogen alone control algal production in Jordan Lake, and pointed out the importance of temperature and light penetration”
(Jackson et al. 1992, p. 20).
Seasonal patterns suggest changing light and temperature may be controlling factors of algal abundance and primary productivity (Kuenzler et al. 1986, p. xiii).
“The major factors controlling phytoplankton growth in nature are usually temperature, light, or one of a few nutrient elements.” (Kuenzler et al. 1986, p. 2).
“Heating not only reduces viscosity, thereby increasing sinking rates of seston (suspended materials, including phytoplankton) but also causes water column stability, reducing the vertical
turbulence which would return seston to the surface (Wetzel 1993).” (Kuenzler et al. 1986, p. 75).
“The seasonal variation in gross productivity and ln gross productivity was strongly correlated with temperature” (Kuenzler et al. 1986, p. 99).
“Linear correlations demonstrated strong positive relationship of gross productivity (g O2 /m2 d) with temperature and negative relationships with light extinction and NO3 concentration.
Turbid waters with high light extinction coefficients caused decreased photosynthetic rates. Periods of high productivity depleted nitrate in the water.” (Kuenzler et al. 1986, p.
104).
“Phytoplankton abundances in B. Everett Jordan Lake appear to be in part limited by factors such as temperature and low light levels, particularly in the cooler months” (Weiss et al.
1985, p. xvii).
“[W]e evaluated correlation between chlorophyll a, both log-transformed and untransformed and a variety of water quality variables (NH3, TIN, OP, TDP, TP, phytoplankton biovolume, Z1%,
water temperatures, and these variables log-transformed.) The results showed chlorophyll a was positively correlated with biovolume and water temperature and negatively correlated
with Z1% and TIN. The other correlations were either not significant or inconsistent.” (Weiss et al. 1986, p. 101).
Jackson et al.’s (1992) modeling results may “indicate that gross primary productivity is not exclusively determined by nutrient levels and phytoplankton biomass. Physical factors such
as turbidity and wind generated mixing may operate to dampen the differences in productivity that would normally be associated with nutrient concentration” (Jackson et al. 1992, p.
20-21).
“… the unusual turbidity in our lakes may keep their trophic status below that predicted from P loading alone (Pearce 1983).” (Kuenzler et al. 1986, p. 91).
Studies on phytoplankton nutrition and growth in Chowan River, N.C. made by Stanley and Hobbie (1977), Sauer and Kuenzler (1981), Kuenzler et al. (1982), Peaerl (1982) … “demonstrated
that either N or P may at times limit algal growth, but that species composition and algal abundance often are controlled by other factors such as light, temperature, salinity, or humic
substances.” (Kuenzler et al. 1986, p. 3-4).

Further studies on blue-green algal growth should include “detailed experimental examination of the roles of inorganic suspended solids on nutrient availability, light extinction, and
algal sedimentation” as well as the roles of hydraulic flushing and inorganic carbon availability (Smith 1987, pp. vii-viii).
Models constructed to predict Jordan Lake’s nutrient/phytoplankton relationships should include parameters that describe its climate, morphology, and hydrology.
It appears that the retention times of both arms is a major factor in establishing net chlorophyll values in relationship to the quantities of nutrients available and the manner in which
these are recycled.” (Weiss et al. 1985, p. xviii).
“Weiss, et al. (1986) … concluded that a useful model of Jordan Lake’s nutrient/phytoplankton relationships would have to incorporate hydrodynamic, climatological, and morphological
variables” (Jackson et al. 1992, p. 20).
“A review of the history of efforts to model Jordan Lake’s primary productivity reveals that the unique hydrological patterns and sensitivity of water movement to localized events may
confound efforts to create a predictive model. The restricted exchange sites between the basins of the New Hope arm, combined with extreme sensitivity of both flow direction and retention
time to inflows results in variable rates of nutrient transport, assimilation, and settling. High levels of turbidity, combined with frequent wind mixing, result in mixing depths which
are frequently deeper than the depth of light penetration. As a result, at certain times of the year, light availability may limit phytoplankton production even when sufficient nutrients
are available” (Jackson et al. 1992, p. 21).
Findings involving the influences of hydrologic events and sedimentation on water quality and algal growth
Hydrologic events greatly impact water quality in Jordan Lake.
Discharges from Haw River, Morgan Creek, New Hope Creek, and Northeast Creek each varied more than two orders of magnitude, resulting in pulse inputs of water, nutrients, and suspended
sediments (Kuenzler et al. 1986, p. 16).
“The distributions of phosphorus fractions in Jordan Lake appeared to be controlled largely by hydrologic events but were also affected by biological processes.” (Kuenzler et al. 1986,
p. 38).
“There was clearly a correlation between Haw River and New Hope River discharges 5 days prior to sampling and light extinction at Stations 5 and 30 [upper New Hope arm and Haw River
arm], respectively” (Kuenzler et al. 1986, p. 77) .

“There were clear correlations (P< 0.01) between stream discharges five days before each sampling trip (Hill et al. 1984; USGS data file) and FRP [filterable reactive phosphorus] concentrations
at Stations 5 and 10” (upper and middle New Hope arms) (Kuenzler et al. 1986, p. 79).
DEM automatic sampler data from Haw River indicates “a large portion of the total annual loads [of phosphorus] enters the reservoir in a relatively few storm events” (Weiss et al. 1985,
p. 34). A two-day event on February 14 and 15, 1984 “accounted for 10% of the estimated annual TP load and transported the equivalent of 61 days of point-source-derived TP” (Weiss
et al. 1986, p. 30).
“The storm load of phosphorus is largely particulate. This rapidly settles from the water column into the deeper waters. Since the storm flow is released from the bottom gates of the
discharge structure, much of the storm-derived phosphorus is almost immediately discharged. This phosphorus, a significant proportion of the annual load, has little impact on the water
column phosphorus concentration.” (Weiss et al. 1986, p. 30).
Sedimentation plays an important role in the Jordan Lake phosphorus cycle.
Occurrence and Importance of Sedimentation
“Nutrient sedimentation down the New Hope arm is facilitated by the limited water flow between basins caused by highway causeways …” (Jackson et al. 1992, p. 11).
“Jordan Lake’s Piedmont watershed supplies abundant suspended sediments, causing lake water to be turbid with suspended clays” (Kuenzler et al. 1986, p. xiii).
“Because clays appear to be carrying phosphate to the bottom of the lake, research seems appropriate to determine the capacity of bottom sediments for long-term storage of P.” (Kuenzler
et al. 1986, p. xviii).
“In the New Hope arm euphotic zone TP concentrations suggest that more than 90% is lost from the inputs. The data suggests, therefore, that the euphotic zone TP concentration is largely
determined by lentic processes (sedimentation / mixing) rather than the input” (Weiss et al. 1985, p. 52).
“Suspended sediments carried into lakes may be important to the P cycle (Hutchinson 1941, Kuenzler and Greer 1980, Jones and Redfield 1984; Cuker 1986), especially when the sediments
are fine clays and have high P binding capacities and long retention times in the water column (Golterman 1973; Syers et al. 1973)” (Kuenzler et al. 1986, p. 7).
Sorption and Bioavailability of Phosphorus in Suspended Sediments
“The algal availability P (AAP) associated with suspended sediments ranged from 5.1 to 18.6% (mean =11.3%) of the total sediment” (Kuenzler et al. 1986, p. 54).

“It appears then, that suspended sediments are not the ultimate source of easily desorbable and available P, but are usually a sink for P during times when P is abundant. High ambient
FRP [filterable reactive phosphorus] concentrations, and often low biomass, immediately after high river flows decrease the importance of sediment-bound P. However, as algal biomass
and photic zone depth increase and dissolved nutrient supplies decrease, the bioavailability of P on the clay particles remaining in suspension becomes more important.” (Kuenzler et
al. 1986, p. 89).
“In spite of several mechanisms by which sediment-associated P can be returned to the water and to the euphotic zone, large amounts of P are lost each year to the lake bottom. Weiss
and Francisco (1985) calculated that 45-55% of TP [total phosphorus] transported by the Haw River and 84-92% of TP transported by the New Hope River were lost in the lake.” (Kuenzler
et al. 1986, p. 91).
Findings involving the segmentation and nutrient gradient in Jordan Lake
The natural segmentation of Jordan Lake makes whole-lake mean values meaningless.
“Whole-lake annual mean values for any biological dimension in B. Everett Jordan Lake are rendered meaningless because segmentation of the lake by road causeways affects flows through
and gradients are quickly established within segments.” Comparing annual means at individual sampling locations appears to be a reasonable procedure (Weiss et al. 1985, p. 53).
The nutrient gradient in the New Hope arm affects plankton production but not gross primary productivity.
There is a nutrient gradient in the New Hope Arm, with the upper portion having much higher nutrient concentrations than the lower section of the arm. Comparisons of inflow and outflows
illustrate “[a]s much as 86 percent of total phosphorus is lost from streams flowing into the New Hope arm, while 48 percent of total nitrogen is lost.” This results in a pronounced
nutrient gradient down the main body of the lake (Jackson et al. 1992, p. 10-11).
Weiss et al. collected monitoring data from 1984-1988. They found that “[m]ean annual values for chlorophyll a, density, and biovolume indicate that plankton production reflects nutrient
availability in its downstream distribution.” Values for middle New Hope arm (Basin III) were generally only 65-70 percent of those of the upper New Hope arm (Basin IV); lower New
Hope arm values (Basin II) ranged from 55-60 percent of the Basin IV values (Jackson et al. 1992, p. 11).

Statistical Analysis of Factors Influencing Eutrophication in Jordan Lake
While past research provides a firm foundation for understanding nutrient response in Jordan Lake, these studies do not reflect more recent data from the 1990’s. This section of the
memo documents how statistical analysis was used as a first step to investigate the correlation between algal response, measured as chlorophyll a concentration, and potential explanatory
variables. As noted in Section 0, the analysis uses the uncorrected chlorophyll a data because of data quality concerns with the corrected chlorophyll a data reported by DWQ. The
presence of correlation does not prove causality; however, the statistical analysis can illuminate the strength of potential predictive relationships as well as validating the conclusions
of earlier researchers.
Potential explanatory variables evaluated in the statistical analysis were: concentrations of various nutrient species (total phosphorus, orthophosphate, total nitrogen, total Kjeldahl
nitrogen, nitrite plus nitrate, total organic nitrogen), total solids, total suspended solids, nonalgal turbidity, Secchi depth, antecedent tributary inflow (5, 7, and 10 day), season
of year, water temperature, and lake segment. The last variable, denoting one of four lake segments, is included as a surrogate for additional unmonitored processes that vary according
to location within the lake. For purposes of the statistical analysis (only) the segments are labeled from the Upper New Hope (Seg = 1) to Haw River Arm/Near Dam (Seg = 4) as shown
in Figure 12. These segments are divided from one another by road causeways or natural constrictions that inhibit exchange. Segment 4 represents the upper end of the New Hope arm,
upstream of SR 1008 (Mt. Carmel Church Road). Segment 3 is the middle part of the New Hope arm, from SR 1008 to the U.S. 64 causeway. Segment 2 is the lower part of the New Hope arm,
from U.S. 64 to the narrows above the Haw River. Finally, Segment 1 represents the Haw River arm and the near dam area. Data available through January 2001 were included in the model.
Model Segmentation As noted by previous researchers, nonalgal turbidity is suspected to be an important control on algal growth. Unfortunately, nonalgal turbidity is not monitored
directly from the lake. Instead, the nonalgal component of turbidity must be inferred from Secchi depth (a measure of water clarity that reflects total turbidity) after correction
for the algal component of turbidity. Initial evaluations used the correction equation proposed by Walker (1987). However, Dodd et al. (1988) noted that this relationship did not
yield accurate values for Jordan Lake, and proposed the following site-specific relationship:

where A is nonalgal turbidity (1/m), Secchi is Secchi depth (m), and Chla is chlorophyll a concentration in µg/L. It should be noted that the estimate of nonalgal turbidity is not independent
of the response variable, chlorophyll a, which may cause statistical tests to over-estimate the significance of nonalgal turbidity as an explanatory variable.
The relationship between uncorrected chlorophyll a and potential explanatory variables was first evaluated using the non-parametric technique of Classification and Regression Trees.
The technique is conceptually simple: Beginning with a single cluster of response variable cases (e.g., chlorophyll a observations), search the candidate set of predictor variables
for a way to split the cluster into two clusters such that the maximum possible reduction in variance is obtained (in this case, using a least squares criterion). Figure 13 shows a
regression tree for Jordan Lake chlorophyll a data. For each step, the distribution of chlorophyll a values is shown, with the central box representing the interquartile range and
the whiskers and isolated points representing values outside this central range. For the Jordan Lake chlorophyll a data, the first split is always made on nonalgal turbidity. In this
case, nonalgal turbidity less than 0.082 m (i.e., high water clarity after correction for algal turbidity) is associated with high chlorophyll a observations. Secondary splits occur
on Lake Segment (SEG) and total phosphorus (TP). None of the other measures under consideration enter the trees as statistically significant explanatory variables.

Classification and Regression Trees Analysis of Chlorophyll a in Jordan Lake

The Regression Tree procedure appears to confirm the important role of nonalgal turbidity in determining chlorophyll a concentrations in Jordan Lake. The strong correlation between
nonalgal turbidity and chlorophyll a concentration is shown in Figure 14. It should be recalled, however, that nonalgal turbidity is not measured independently of chlorophyll a, so
the apparent strength of this relationship may be misleading. Similarly, total organic N is also affected by the nitrogen component of algal cells in water quality samples.

Relationship of Chlorophyll a Concentration to Non-Algal Turbidity

The statistical investigation was continued by applying stepwise multiple linear regression to the data set. To reduce problems with heteroscedasticity (scale-dependent variance), chlorophyll
a, nonalgal turbidity, and nutrient concentrations were converted to a natural logarithm basis. The stepwise regression procedure identified five of the candidate variables as statistically
significant, and leads to the following equation:

in which UCHL is uncorrected chlorophyll a ((g/L), NAT is nonalgal turbidity (1/m) by Dodd’s method, TON is total organic nitrogen (mg/L), TP is total phosphorus (mg/L), Seg is segment
number, and Temp is water temperature (Celsius). This regression model appears to have good explanatory power, with an adjusted multiple R2 of 83.6% and a standard error of the estimate
of 0.348. All of the model coefficients are statistically significant at the 95 percent level. Observations versus predictions of the model are shown in Figure 15.
In this model, the coefficient on TON is negative, even though the correlation between chlorophyll a and TON is positive. This occurs because TON is positively correlated with TP and
negatively correlated with Seg. The positive correlation between chlorophyll a and TON is explained by these other variables.

Multiple Regression Model of Uncorrected Chlorophyll a in Jordan Lake.

Algal Community Ecology
The status of algae in Jordan Lake is most often thought of in terms of the concentration of chlorophyll a, which is the main photosynthetic pigment in most algae. But, chlorophyll
a is only an imprecise indicator of actual algal density, as the chlorophyll a content varies significantly among species and with light intensity and nutrient availability. It also
provides no indication of whether undesirable species predominate. This section of the memo summarizes additional insights gained into lake response by examining data on algal biovolume
and abundance of specific taxa.
Algal biovolume is a direct measure of the volume of a sample taken up by algae, and is usually expressed as mm3/m3. Algal biovolume relates directly to some impairments of use, e.g.,
unaesthetic conditions. DWQ has reported a large number of algal biovolume analyses for Jordan Lake from January 1983 through October 1989. This period was selected for detailed analysis
because biovolume data were collected much more frequently prior to 1990 than in subsequent years. Biovolumes are obtained by identifying the algal taxa present, counting the cells
or units, and converting to a biovolume basis through use of a database of typical volumes for each taxon.
Algal biovolume generally increases in spring and summer, decreasing in the fall and winter (Figure 16 and Figure 17). Biovolume appears to exhibit an inverse relationship to nitrate
plus nitrite (NOx) concentrations. Nutrients like NOx can become depleted in the photic zone during the summer due to uptake by algae.

Algal Biovolume and NOx Concentrations in New Hope Arm of
Jordan Lake (Station CPF087B3), 1983 - 1989.

Algal Biovolume and NOx Concentrations in Haw River Arm of
Jordan Lake (Station CPF055C), 1983 - 1989.
The New Hope arm has generally higher and less variable biovolumes than Haw River arm. The New Hope arm also has more prolonged loss of NOx in summer than the Haw River arm. This reflects
the fact that the Haw River arm has a much shorter residence time and is more likely to exhibit rapid changes in response to changing inflows.
Algal group (division) trends in biovolume are shown for three representative dates in Figure 18 through Figure 20. The order of the algal groups in the stacked bar charts, from top
to bottom, are:
yellow-green algae [chr = chrysophytes]
prymnesioids [pry]
euglenoids [eug]
blue-green algae [cya = cyanophytes]
cryptomonads [cry]
green algae [chl = chlorophytes]
diatoms [bac = bacillariophytes]

In winter, the diatoms tend to be the most abundant algal group in Jordan Lake. In summer, the blue-green algae tend to be most abundant. While these patterns are typical of Jordan
Lake, patterns of abundance in all seasons shift from year to year. Indeed, many temporal and spatial changes in algal abundance may be random, or at least difficult to explain with
water quality data.

Algal Biovolume Distribution, February 26, 1987

Algal Biovolume Distribution, August 25, 1987

Algal Biovolume Distribution, August 1, 1989
Certain algal taxa are of particular concern to water supply managers due to the potential for filter clogging or taste and odor problems in finished water. Nuisance characteristics
of dominant algal taxa found in Jordan Lake are summarized in Table 5. The columns labeled “Eutrophication” and “Pollution Tolerance” denote taxa that are indicators of nuisance conditions
associated with nutrient enrichment or general pollution. The columns labeled “Taste & Odor” and “Filter-clogging” indicate specific nuisance conditions potentially caused by the taxon.
Of the 24 most prevalent algae identified in Jordan Lake, 18 are associated with at least one nuisance condition. However, based on the available monitoring data, the most notorious
bloom-causing and toxin-producing algal species are not dominant in the lake.
Nuisance Associations of Dominant Algal Taxa Found in Jordan Lake
Division Genus Species Variation Eutrophi-cation Pollution Tolerance Taste & Odor Filter-clogging CYA Anabaena sphaerica X CYA Anabaena spiroides X CYA Anabaena X X X CYA Anacystis cya
nea CHL Ankistrodesmus falcatus spiriliformis X CHL Chlamydomonas globosa X CRY Chroomonas caudata PRY Chrysochromulina breviturrita CHR Chrysococcus minutus CRY Cryp
tomonas erosa X X CRY Cryptomonas ovata X CHR Mallomonas caudata X BAC Melosira granulata angustissima X X X X BAC Melosira italica alpigena X X BAC Melosira italica tenuissima X X
BAC Melosira italica X X CHR Ochromonas species CYA Oscillatoria geminata X CYA Phormidium angustissimum X CHL Scenedesmus acuminatus X X CHL Tetraedron trigonum X EUG
Trachelomonas crebea obesa X EUG Trachelomonas hispida punctata EUG Trachelomonas volvocina X
Examination of the prevalence of specific algal taxa also reveals some additional information on the general relationships controlling algal biovolume. In particular, the biovolume
of blue-green algal taxa is correlated with water temperature (positive) and NOx (negative), as shown in Figure 21 and Figure 22. These relationships make sense because blue-greens
are most abundant in summer, when the water is warm and NOx is depleted. In addition, some blue-greens are able to extract or fix nitrogen from the atmosphere, allowing them to grow
and dominate when nitrogen availability limits growth of other algae. For more detailed statistical analysis of algal community data, see Appendix C.

Correlation of Blue-Green Algal Biovolume and Temperature
(sample dates: 2/83 – 10/89, 6/96 – 9/97, 4/99 - 8/99)

Correlation of Blue-Green Algal Biovolume and NOx Concentration
(sample dates: 2/83 – 10/89, 6/96 – 9/97, 4/99 - 8/99)

Scoping-Level Model of Lake Nutrient Response
As part of the evaluation of existing data, Tetra Tech developed and applied a simplified scoping model of lake response. Use of such a scoping model yields further insights on existing
data, provides initial information to guide the final modeling effort, and gives an indication of potential modeling results. This section of the memo documents the results of Tetra
Tech’s scoping model analysis.
The scoping model application uses the U.S. Army Corps of Engineers’ BATHTUB model (Walker, 1987). The BATHTUB model is designed to facilitate application of empirical eutrophication
models to morphometrically complex reservoirs. The program performs water and nutrient balance calculations in a steady-state, spatially segmented hydraulic network that accounts for
advective transport, diffusion, and nutrient sedimentation. Eutrophication-related water quality conditions are expressed in terms of total phosphorus, total nitrogen, chlorophyll
a, transparency, organic nitrogen, non-orthophosphate phosphorus, and hypolimnetic oxygen depletion rate. These conditions are predicted using semi-empirical relationships developed
and tested on a wide range of reservoirs (Walker 1985). The model can also simulate mass balances of arbitrary conservative substances (such as chloride).
Mass balances are computed in BATHTUB at steady state over an appropriate averaging period. Steady-state approximation means that only seasonal or annual average loads and lake conditions
are simulated, although the loads and conditions may change from year to year. In other words, the model does not represent day-to-day changes in flow, loads, or nutrient concentrations.
Although this approach represents a compromise, it has proven effective in practice. Shortterm variations in lake conditions reflect variations in flow, including wind and weather
effects, which require complex and labor-intensive models; such effects tend to average out, however, over longer time frames. Thus, annual or seasonal average conditions can be successfully
predicted using data that are insufficient for simulating day-to-day variability.
While BATHTUB takes a steady-state approach to variability in time, it can represent spatial complexity. Within the model, a reservoir can be represented as a set of linked segments,
with as much detail as desired. Each segment is assumed to be laterally mixed, but the water column can be represented as stratified into three layers (each of which is assumed to
be of constant thickness over the averaging period, in accordance with the model’s steady-state assumptions). Exchanges of constituents between adjacent segments can occur by advection
and dispersive transport, both of which are calculated by the model. The model includes several methods for representing dispersive exchange.
BATHTUB also provides a variety of options for simulating nutrient sedimentation, including several first- and second-order representations proposed in the literature, as well as methods
developed explicitly for BATHTUB. Also available are five submodels for chlorophyll a, which depend variously on nitrogen, phosphorus, light, and flushing rate limitations, and three
candidate models relating Secchi depth (transparency) to chlorophyll a, turbidity, and nutrient concentrations. BATHTUB thus provides a highly flexible tool for developing a semi-empirical,
annual-average analysis of nutrient concentrations and eutrophication. Excellent results have been obtained in previous applications of the BATHTUB model to both Falls Lake and Cane
Creek Reservoir (Butcher et al., 1995; Butcher et al., 1996).
Model Segmentation
For application of the BATHTUB model, Jordan Lake was divided into four segments, as described previously and shown in Figure 12. Water quality monitoring data are available to support
model calibration for each of these segments. Further refinement of segmentation will likely be needed for the ultimate nutrient response model, but the four-segment representation
is adequate for scoping.
Inflows to the lake occur predominantly to Segment 1 (Haw River arm) and Segment 4 (Upper New Hope, receiving flow from New Hope Creek, Morgan Creek, and Northeast Creek among others).
Other smaller inflows are also accounted for in the model. A schematic representation of the modeling network is shown in Figure 23.

Schematic Representation of BATHTUB Model Network for Jordan Lake

Specification of External Loads
BATHTUB requires specification of the flows and loads or flow-weighted concentrations of nutrients in tributaries to each segment. For the four largest tributaries (Haw River, New Hope
Creek, Morgan Creek, and Northeast Creek), continuous flow monitoring and frequent point-in-time concentration data are available.
Estimating constituent mass loads from point-in-time measurements of water-column concentrations presents many difficulties. Load is determined from concentration multiplied by flow,
and while measurements of flow are continuous (daily average), only intermittent (e.g., monthly or tri-weekly grab) measurements of concentration are available. Calculating total load
therefore requires "filling in" concentration estimates for days without samples. The process is further complicated by the fact that concentration and flow are often highly correlated
with one another, and many different types of correlation may apply. For instance, if a load occurs primarily as a result of nonpoint soil erosion, flow and concentration will tend
to be positively correlated; that is, concentrations will increase during high flows, which correspond to precipitation-washoff events. On the other hand, if load is attributable to
a relatively constant point discharge, concentration will decrease as additional flow dilutes the constant load. In most cases, a combination of processes is found.
Preston et al. (1989) undertook a detailed study of advantages and disadvantages of various methods for calculating load from tributary concentration and flow data. Their study demonstrates
that simply calculating load for days when both flow and concentration have been measured and using results as a basis for averaging is seldom a good choice. Depending on the nature
of the relationship between flow and concentration, more reliable results may be obtained by one of three approaches:
Averaging Methods: An average (e.g., yearly, seasonal, or monthly) concentration value is combined with the complete time series of daily average flows;
Regression Methods: A linear, log-linear, or exponential relationship is assumed to hold between concentration and flow, thus yielding a rating-curve approach; and
Ratio Methods: Adapted from sampling theory, load estimates by this method are based on the flow-weighted average concentration times the mean flow over the averaging period and performs
best when flow and concentration are only weakly related.
No single method provided superior results in all cases examined by Preston et al.; the best method for extrapolating from limited sample data depends on the nature of the relationship
between flow and concentration, which is typically not known in detail. Preston et al. show that stratification of the sample, however, almost always reduces error in estimation.
Stratification refers to dividing the sample into two or more parts, each of which is analyzed separately to determine the relationship between flow, concentration, and load. Sample
data are usually stratified into high- and low-flow portions, allowing a different relationship between flow and load at lowflow (e.g., diluting a constant base load) and highflow
regimes (e.g., increasing load and flow during nonpoint washoff events). Stratification could also be based on time or season to account for temporal or seasonal changes in loading.
Selection of an appropriate statistical method and stratification scheme for estimating tributary load of a particular constituent can be a difficult and time-consuming process. To
assist in the analysis of annual or seasonal loads within a watershed above a given monitoring station, the COE Waterways Experiment Station developed the FLUX modeling package (Walker
1987 ). FLUX contains options for six calculation methods (one averaging, two ratio, and three regression methods), with flexible options for sample stratification. FLUX also provides
extensive capabilities for analysis of goodness-of-fit of results and other diagnostics and is designed to provide input to the COE lake eutrophication model, BATHTUB, which is also
used as an initial scoping model in this study.
To obtain an estimate of the error coefficient of variation (CV) for each load estimation method, FLUX uses the jackknife procedure, which excludes each measured concentration value,
one at a time, and recalculates loadings (Mosteller and Tukey 1978 ). Error variance is then estimated empirically based on the difference between mean flux calculated with and without
each individual sample to compare calculation methods on a uniform basis. As demonstrated by Walker, the jackknife procedure also yields a reasonably unbiased estimate of the actual
error variance in most cases (1987). The CV thus provides a basis for evaluating different load estimation methods, but does not necessarily accurately estimate the true error variance
associated with an estimate, because (1) these variance estimates do not account for errors introduced by unrepresentative samples, particularly in the sampled flow distribution, and
(2) the variance estimates do not reflect the bias associated with some calculation methods under certain conditions. Accordingly, the CV is a primary factor in selecting a load estimation
method, but sample characteristics and bias potential must also be considered before making the final selection.
The FLUX model was used to estimate annual flow-weighted concentrations for nutrients on an annual basis. Complete details of the methods used and results are provided in Appendix D.
There is also a portion of the watershed draining to the lake that is not monitored (refer to map, Figure 6-3). This consists primarily of the nearshore drainage that is either downstream
of the monitoring stations or drained by smaller streams, although Little Creek, which drains the northern part of Chapel Hill is also unmonitored. The unmonitored area constitutes
12.6 percent of the total area draining to the lake.
To estimate loads from the unmonitored portion of the Jordan Lake Watershed the area was delineated into four separate areas corresponding to the drainage areas of the major tributaries
entering the lake (refer to map in Figure 24). Within ArcView, the delineated watershed boundaries were imposed on unfiltered 1996 EarthSat land use/land cover data provided by the
Triangle-J Council of Governments. The land use/land cover classifications defined in the EarthSat data were then aggregated into a broader land use classification scheme corresponding
to established export coefficients which could be utilized to estimate nutrient loads from the unmonitored area based on rainfall and runoff rates.
Nonpoint loading from the unmonitored area was estimated using a simple export coefficient approach with modification to account for year to year variability in precipitation. The general
approach is similar to the SIMPLE model of Schueler (Caraco et al., 1998) as augmented for application to the Randleman Lake study (Tetra Tech, 1998). In this approach, runoff is first
divided into surface and baseflow components. The surface component is estimated as
Rs = 0.95 x P x I + 0.11 x P x (1-I)
where Rs is the surface runoff depth, P is the annual precipitation depth, and I is the impervious surface fraction. The coefficient of 0.11 on the pervious runoff is larger than the
SIMPLE model default of 0.05, based on analysis of flow records in the Deep River area.
The baseflow portion of runoff, Rb, is estimated as a fraction of precipitation on pervious surfaces with a minimum depth:
Rb = min {7.5”, 0.5 x P x (1-I)}
The total annual runoff depth is then Rt = Rs + Rb.
The total average runoff concentration in a given year, Ct, arises as a weighted average of the concentration in surface runoff, Cs, and the concentration in baseflow, Cb. This yields
Ct = Rs/Rt x (Cs – Cb) + Cb
Typical values for concentration in runoff from the literature are assigned as shown in Table 6. Separate concentration values are used for developed/managed areas (including urban,
residential, and agricultural uses) and undeveloped areas (vacant, forest).
Nonpoint Source Runoff Nutrient Concentrations Assumed
for Unmonitored Areas of Jordan Lake Watershed (in mg/l)
Total Phosphorus Total Nitrogen Cs – Developed/managed areas 0.225 2.00 Cs – Undeveloped areas 0.100 1.40 Cb (baseflow concentration) 0.035 1.00
This simple approach allows long-term average runoff concentrations to be set equal to literature values while providing a representation of year-to-year variability associated with
precipitation experienced in a given year.
For input to the BATHTUB model, the calculated runoff and concentration is used to estimate total nonpoint nutrient loads for each unmonitored area. Total phosphorus and nitrogen were
partitioned to organic and inorganic fractions assuming a 50:50 mix. This is a reasonable approximation based on analysis of stream data at Falls Lake and Cane Creek Reservoir (Butcher
et al., 1995, 1996).

Map of Unmonitored Portion of Jordan Lake Watershed
Finally, the Pittsboro WWTP also discharges to Roberson Creek, which is part of the unmonitored area draining directly to the Haw River Arm of the lake. For the purpose of this scoping
model, the additional loading from effluent from this plant was assumed to pass to the lake without reduction in transit, given the short distance and travel time to the lake. Loading
estimates for the Pittsboro WWTP were obtained from compliance data for the specific years simulated in the BATHTUB model. The WWTP phosphorus load was assumed to be 75 percent inorganic.
Nutrient loads from several small privately-owned wastewater treatment facilities located in the unmonitored watershed area were not estimated for this scoping-level analysis.
Nutrient loads from both the monitored and unmonitored drainage areas are summarized for 1997, one of the model calibration years, in Table 7. Relative contributions of different sources
are summarized in Figure 25 and Figure 26.
Estimated Nutrient Loads for Jordan Lake BATHTUB Analysis, 1997
NUTRIENT LOADS FROM FLUX ANALYSIS Mean Flow
(m3/yr * 106) [mgd] Total P
Avg. Conc.
(ppb) Total N
Avg. Conc.
(ppb) Total P
Mass Load
(kg/yr) Total N
Mass Load
(kg/yr) Haw River 1,007.1 [728.9] 147 1,138 148,040 1,146,080 Morgan Creek 34.7 [25.1] 138 2,619 4,790 90,930 New Hope Creek 96.1 [69.6] 174 2,187 16,720 210,170 Northeast Creek 24.1
[17.4] 332 3,357 7,990 80,840 OTHER NUTRIENT LOADS
(LU/LC Analysis) Mean Flow
(m3/yr * 106) [mgd] Total P
Avg. Conc.
(ppb) Total N
Avg. Conc.
(ppb) Total P
Mass Load
(kg/yr) Total N
Mass Load
(kg/yr) Haw River nps 56.3 [40.7] 38 712 2,140 40,050 Lower New Hope nps 40.9 [29.6] 10 715 410 29,210 Middle New Hope nps 38.5 [27.9] 39 709 1,500 27,300 Upper New Hope nps 74.6
[54.0] 40 696 2,980 51,920 WASTEWATER TREATMENT PLANT WITHOUT FLUX ANALYSIS Mean Flow
(m3/yr * 106) [mgd] Total P
Avg. Conc.
(ppb) Total N
Avg. Conc.
(ppb) Total P
Mass Load
(kg/yr) Total N
Mass Load
(kg/yr) Pittsboro WWTP 0.5 [0.4] 1,712 10,316 800 4,850

Estimated Nitrogen Loading to Jordan Lake, 1997

Estimated Phosphorus Loading to Jordan Lake, 1997
BATHTUB Calibration
Calendar year 1997 was initially selected to begin calibration of the Jordan Lake BATHTUB scoping model because it was (a) the most recent year of record for which monitoring data were
complete and available for all in-lake stations and (b) not subject to significant climatic anomalies due to hurricanes passing near or through the watershed. To find additional years
to utilize for model calibration and verification, it was necessary to revert back to 1989 and 1990 to obtain years with thorough data and no hurricane disturbances. As a steady-state
average model, BATHTUB is not expected to provide accurate results for years in which a significant portion of the annual inflow is associated with a hurricane or other extreme hydrologic
events. Initial efforts revealed some anomalies with the 1997 data. In particular, most tributary data for the second half of this year are missing from the DWQ files, so external
load estimates are uncertain. As a result, 1990 was substituted as the primary calibration year and 1997 and 1989 were used for model testing. Appendix E contains complete BATHTUB
input files for all years utilized in this scoping model analysis.
The first step in application of BATHTUB to a multiple-segment lake is specification of rates of mixing or dispersion between segments. Dispersion may be calibrated if a conservative
tracer is available. There are not sufficient spatial gradients in chloride concentrations across the lake to use as a tracer. Specific conductance values do show a spatial gradient.
While this is not a fully conservative tracer, segment-to-segment gradients in specific conductance do provide qualitative information on rates of inter-segment mixing. The specific
conductance data from 1997 were therefore used to estimate average rates of dispersion between segments. As expected from the morphometry of the lake, dispersion coefficients between
segments are non-zero but well less than the default BATHTUB values. Dispersion calibration factors that produced a good fit to the 1997 data (shown in Table 8) were then tested on
calendar year 1990. The selected coefficients produced a plus or minus 10% error in specific conductance for the 1990-year scenario, when compared to observed data for each segment,
and were deemed appropriate for model use.
Through evaluation of nutrient residence time and turnover ratios as per BATHTUB guidance (Walker, 1996), it was determined that one year was the optimum simulation period for Jordan
Lake inflows. Shorter residence times in the Haw River Arm suggest that a seasonal model would be more appropriate in that section of the lake; however, the same simulation period
must be used for all segments of the model.
The primary calibration procedure for BATHTUB consists of adjusting nutrient sedimentation rates to match observed data. These semi-empirical sedimentation rates account for the net
change in concentration within a lake segment due to all processes, including deposition, regeneration from the sediments, and, in the case of nitrogen, exchanges of gas with the atmosphere.
Phosphorus loss was simulated using BATHTUB Model 2, a second-order decay rate function in which the unit phosphorus sedimentation rate (mg/m3-yr) is calculated as

where P is the total phosphorus concentration (µg/L), Fot is the fraction of ortho-phosphate in the tributary total phosphorus load, Qs is the segment overflow rate (m/yr), and CP is
the phosphorus sedimentation calibration factor. Based on successful application to Falls Lake, nitrogen losses were simulated using BATHTUB Model 7, a simple first-order settling
model in which the unit nitrogen sedimentation rate (mg/m3-yr) is given by

where N is the total nitrogen concentration (µg/L), Z is the segment average depth (m), and CN is the nitrogen calibration factor.
Given that the three segments that comprise the New Hope arm of the reservoir (Segments 2-4) have similar physical and hydrodynamic characteristics they were initially grouped together
for purposes of calibrating nitrogen and phosphorus settling rates. The Haw River arm (Segment 1) was treated a separate and distinct segment group. BATHTUB was run to solve for phosphorus
and nitrogen settling rates for the two segment groups that produced the optimum simulation of observed growing season average chlorophyll a concentrations for 1990, which were then
tested on the 1997 data. Slight adjustments were made to the settling rates to produce the best compromise fit between 1997 and 1990 and the compromise rates were then tested on the
1989 data for validation. The calibrated sedimentation rates resulted in a reasonable fit to observed 1989 segment-averaged nutrient data and were deemed appropriate. Finally, calibration
factors were assigned to chlorophyll a response, using the BATHTUB model option that includes potential co-limitation by nitrogen, phosphorus, and light availability.
Table 8 lists the values for each of the model calibration factors. To reproduce observed gradients in the lake it was necessary to assign lower coefficients to the upper New Hope arm
and higher nutrient sedimentation rates to the Haw River arm. High rates of sedimentation in the Haw River arm relative to tributary monitoring are appropriate because of the sharp
change in velocity between the free-flowing Haw River and the impoundment.
Calibration Factors for Jordan Lake BATHTUB Model
Segment Number Segment Name Phosphorus Sedimentation Factor, CP Nitrogen Sedimentation Factor, CN Chlorophyll a Calibration Factor Dispersion
Calibration Factor 1. Haw River 4.00 97.5 1.50 NA 2. Lower New Hope 2.00 19.7 2.48 0.1 3. Middle New Hope 2.00 19.7 2.48 0.6 4. Upper New Hope 0.30 19.7 2.03 0.2
Figure 27 shows the nutrient calibration results. While the fit is generally good, the plot reveals some discrepancies between the model and data. For phosphorus, the model over-estimates
concentrations in the Lower New Hope segment in 1990, while underestimating concentrations in the Upper New Hope in 1989. Nitrogen is under-estimated in the Middle New Hope segment
in 1990 and over-estimated in the Upper New Hope segment in 1997. To some extent, these discrepancies may reflect the quality of the data (limited number of in-lake samples to estimate
averages, missing tributary data for the second half of 1997). Discrepancies may also arise from the simple steady-state nature of the model, which may not be adequate to represent
the effects of transient mixing events between the New Hope and Haw River arms of the lake.
Figure 28 shows the chlorophyll a calibration results, with the model calibrated or “tuned” to the 1990 data. Over-predictions of chlorophyll a concentrations in 1997 in the Upper and
Lower New Hope segments mirror the over-predictions of nitrogen and phosphorus concentrations in these segments respectively. Further, as with the nutrients, the model is not intended
to, nor is it capable of reproducing intra-season variability in chlorophyll a, such as might arise in response to flushing events or temporal changes in non-algal turbidity and light
availability.
The BATHTUB calibration effort led to the following conclusions:
Use of a unified parameter set generates results that are “fair” at best for nutrient concentrations and chlorophyll a across multiple years. This may suggest that there are differences
in nutrient loss and algal responses between years, reflecting intra-year hydrologic patterns.
The data available for calibration to chlorophyll a are limited (generally monthly) with respect to the natural variability inherent in algal responses. Thus the calibration targets
may not accurately represent true growing season averages.
Chlorophyll a and nutrient concentration gradients in the Lower New Hope segment are the most difficult to fit. This could reflect the influence of backflow exchange from the Haw River
arm into this segment.
Because the temporal variability in algal response is not modeled by BATHTUB, responses to turbidity pulses cannot be represented.

BATHTUB Nutrient Calibration for Jordan Lake

BATHTUB Summer Average Chlorophyll a Calibration for Jordan Lake
BATHTUB Model Analysis of Sensitivity of Lake Response to Changing Nutrient Loads
Despite its simplicity and inherent limitations, the BATHTUB model provides a useful scoping tool for initial investigations of the potential sensitivity of lake response to changes
in nutrient loads. The response variable for the sensitivity analysis is predicted summer average chlorophyll a concentration. An analysis of chlorophyll a response to changes in
nutrient loads was conducted by individually varying nitrogen and phosphorus loads by factors of plus or minus 25 percent. Nutrient load inputs were varied for the Haw River first and
then separately for all tributaries to the New Hope arm of the reservoir in order to evaluate the variation in response caused by changing nutrient loads to the different arms. Finally,
for each nutrient, loads were varied simultaneously for all tributaries by plus or minus 25 percent.
Results for the analysis of sensitivity to varying nitrogen loads are shown in Figure 29 and Figure 30. Figure 29a shows that increasing or decreasing the nitrogen load from the Haw
River by 25% results in an increase of 16 percent and a decrease of 22 percent in predicted chlorophyll a concentrations within the Haw River Segment, respectively. Smaller changes
also occur up the New Hope Arm due to mixing from the Haw River arm, but there is little effect in the Upper New Hope segment. Figure 29b shows that varying the nitrogen loads in New
Hope Arm tributaries by plus or minus 25 percent produces predicted chlorophyll a concentration increases of 9 to 19 percent and decreases of 10 to 22 percent in New Hope segments,
respectively. Almost no change in chlorophyll a is predicted for the Haw River segment. Varying the nitrogen loads in all tributaries together, while keeping the phosphorus loads
the same, results in changes in chlorophyll a concentrations similar to the sum of changes predicted from varying loads to the separate arms individually (Figure 29c). The nitrogen
load analysis indicates that the chlorophyll a level in the reservoir is estimated to be sensitive to changes in nitrogen load inputs, and that the degree of sensitivity to reductions
may be somewhat greater than the degree of sensitivity to increases.
Figure 30a through Figure 30c show the chlorophyll a changes predicted to result from variation of tributary phosphorus loads while maintaining nitrogen loads at current levels. It
is important to note that the y-axis scale, representing change in chlorophyll a, in these three figures is half of the scale shown in Figure 29 for the nitrogen analysis. Figure 30
shows that the predicted responses to 25 percent variations in tributary phosphorus loads are much smaller than those predicted to result from the same percentage changes in tributary
nitrogen loads. The largest chlorophyll a increase predicted to result from a 25 percent increase in phosphorus loading in any segment is 3.4 percent, which occurs in the Upper New
Hope segment. The largest decrease that is predicted to occur as a result of decreasing phosphorus is –5.0 percent, also in the Upper New Hope. As with the nitrogen analysis, the
phosphorus load sensitivity analysis suggests that the predicted chlorophyll a response to phosphorus load decreases was greater than that for phosphorus load increases.

Sensitivity of Jordan Lake Summer Average Chlorophyll a Concentrations to 25 Percent Changes in Tributary Nitrogen Loads (BATHTUB Analysis Based on 1997 Conditions).

Sensitivity of Jordan Lake Summer Average Chlorophyll a Concentrations to 25 Percent Changes in Tributary Phosphorus Loads (BATHTUB Analysis Based on 1997 Conditions).
Initial Review of In-Lake Water Quality Monitoring Program Framework
As summarized in Section 2, several agencies, including DWQ, the USGS, and a number of local governments, conduct water quality monitoring in or immediately adjacent to Jordan Lake.
These monitoring activities have been conducted under different programs, and are not part of a comprehensive or systematic approach to water quality monitoring and data interpretation
for Jordan Lake.
Table 9 summarizes Jordan lake monitoring efforts, including the parameters monitored, the frequency of monitoring, and the parties conducting the monitoring at Jordan Lake. The following
initial observations are made:
DWQ has recently temporarily increased the frequency of in-lake monitoring; however, there is a limited amount of information concerning growing season conditions in the lake.
there is a limited amount of information concerning algal species, counts, and biovolume;
algal biomass is not monitored in Jordan Lake;
corrected chlorophyll a values are only available to September 1996 due to problems with equipment and calibration; rather than omit the data from the most recent years, we decided to
use the uncorrected values for the trend analyses;
there is only limited information relating to sediment conditions and regeneration from sediments;
there is no monitoring of water flow and mixing between the lake’s segments;
continuous (intra-day) monitoring studies to support calibration of the diurnal DO cycle are not available;
there are not sufficient spatial gradients in chloride concentrations across the lake to use as a tracer for model development;
there is limited spatial coverage. The need for additional in-lake and near-shore tributary monitoring sites should be considered;
there is limited data regarding pollutant loads under various hydrologic conditions;
the lack of nutrient samples from the lake’s bottom water layer limits model calibration and verification;
DWQ’s ambient water quality monitoring data for sites located throughout the lake’s watershed may need further refinement to include certain nutrient parameters.
Summary of Nutrient Response Related Water Quality Monitoring Efforts in Jordan Lake, Showing Monitoring Agency, Parametric Coverage, and Approximate Sampling Frequency
AGENCY # Lake Sites # Near-Lake Sites TP PO4 NH3N NOx TN Chl. A Algal Growth DO pH Type Biomass Biovolume DWQ
13 4 ( ( ( ( ( ( ( ( ( ( USGS/ TAWSMP 4 3 ( ( ( ( ( ( Cary WTP 1 0 Cary WTP only monitors three types of algae ( (
Based on the above, the following initial recommendations are offered for consideration by the Local Government Project Partners, water systems using Jordan Lake as a water supply source,
other interested local governments, and State and Federal agencies. Additional recommendations may be included in the final project report.
a comprehensive, systematic, and coordinated Jordan Lake water quality monitoring, research, and data interpretation and analysis program should be developed and implemented, such as
has been done for the Occoquan Reservoir, the San Francisco Institute, and other cooperative monitoring programs;
water quality monitoring should address parameters of concern from a drinking water supply perspective, recreational use perspective, and aquatic protection perspective;
the frequency of monitoring should be designed to adequately assess current conditions and trends, and to support the calibration, verification, and refinement of water quality models;
the sampling program should be developed with input from all of the above parties, as well as university researchers and other appropriate parties.

Summary and Conclusion
Jordan Lake is eutrophic, with high algal productivity, especially in the upstream ends of the New Hope and Haw River arms. Conditions in the lake appear to have improved somewhat from
lake startup until the early 1990’s, but have shown little change since that time.
Key results of the evaluation of existing data include the following:
Excessive algal growth in the lake is supported by high levels of nutrient input and recycling.
Several lines of evidence, including an initial BATHTUB scoping model, N:P ratio, and algal growth potential tests, suggest that algal response in the lake is sensitive to nitrogen loading,
with less sensitivity to phosphorus loading.
Nonalgal turbidity and consequent reduction in light penetration plays an important role in controlling algal growth in the lake.
Nutrient cycling in the lake is complex, with strong feedback between algal populations and concentrations of nutrient species. Algal biomass is sufficiently high that dissolved inorganic
forms of nutrients are rapidly scavenged from the water column during the growing season.
Algal response to nutrient loads differs among taxonomic groups. In particular, blue green algae (Cyanophytes) have a high correlation with summer nuisance conditions, but show a negative
correlation with nitrate-plus-nitrite nitrogen concentration.
Because of the complex responses of different algal groups to nutrient loads, chlorophyll a concentrations provide only an approximate and rough indicator of responses that may degrade
or impair uses of the lake.
Mixing patterns in the lake are complex, and involve exchanges between the Haw River and New Hope arms. These two segments have very different hydraulic characteristics and residence
times, and may exhibit qualitatively different responses to changes in nutrient loads.
This memorandum also documents the application of a scoping-level, steady-state model of average annual lake response. This model does an adequate job of explaining the spatial gradients
in average chlorophyll a concentrations in the lake in a given year. A single set of model parameters does not, however, appear to fully capture the year-to-year variability in lake
response. This likely reflects: (1) variability in nutrient loss to sedimentation associated with differing hydraulic patterns, (2) the complex, time-dependent interaction between
the New Hope and Haw River segments, and (3) the inability of the model to fully distinguish between algal responses in the short residence time Haw River arm versus the longer residence
time New Hope arm.
Accordingly, it appears that a more sophisticated modeling approach is required to meet the needs for a calibrated nutrient response model, as specified in NC General Statute 143- 215.1.
Specifically, an appropriate deterministic modeling tool should be able to:
address dynamic changes in response on an intra-seasonal scale;
represent the actual pattern of mixing between lake segments; and
include a representation of nutrient cycling that can represent nutrient-algal and water column-sediment interactions at a more sophisticated, process-based level.

References
APHA (1985) Standard Methods for the Examination of Water and Wastewater, 16th Edition. American Public Health Association, Washington, DC.
Baker, K.S., Smith, R.C., Nelson, J.R. (1983) Chlorophyll Determinations with Filter Fluorometer: Lamp/Filter Combination Can Minimize Error. Limnology and Oceanography, Volume 28, Issue
5, 1037-1040.
Butcher, J., T. Clements, A. Beach, K. Brewer, D. Korn, N. Archambault, and P. Kellar. 1995. Falls Lake Watershed Study—Final Report. Prepared for North Carolina Department of Environment,
Health, and Natural Resources. The Cadmus Group, Durham, NC.
Butcher, J., T. Clements, K. Brewer, A. Werner, D. Korn, J. Carey, N. Archambault, S. Coffey, D. Line, and G. Pesacreta. 1996. Cane Creek Reservoir Watershed Study. Report to Orange
Water and Sewer Authority, Carrboro, NC. The Cadmus Group, Durham, NC.
Caraco, D., R. Claytor, and J. Zielinski. 1998. Nutrient Loading from Conventional and Innovative Site Development. The Center for Watershed Protection, Ellicott City, MD.
Dodd, R.C., J.F. Smith, and J.D. Vogt. 1988. The development of a phosphorus management strategy for two Piedmont reservoirs in North Carolina. Lake and Reservoir Management, 4(2):
243-252.
Jackson, J. R., Rice, R. A., Nobel, R. L., and Mozley, S. C. Mechanisms of Reservoir Fish Community Dynamics. [Jordan Lake]. Federal Aid in Fish Restoration Project F-30-1. North
Carolina Wildlife Resource Commission, Division of Boating and Inland Fisheries. 1992.
Hoyer, M.V. and Jones, J. R. “Factors Affecting the Relation Between Phosphorus and Chlorophyll a in Midwestern Reservoirs.” Canadian Journal of Fisheries and Aquatic Sciences, 40
(2): 192-199. 1983.
Kuenzler, E.J., Belense, A. J., and Rudek, J. Nutrient Cycling and Productivity of a North Carolina Piedmont Reservoir. WWRI Report No. 228. Water Resources Research Institute of
the University of North Carolina, Raleigh, N.C. 1986.
NCDEM. 1992. North Carolina Lake Assessment Report. Report No. 92-02. North Carolina Department of Environment, Health, and Natural Resources, Division of Environmental Management,
Water Quality Section. June 1992.
North Carolina Division of Water Quality. Water Quality Conditions B. Everett Jordan Reservoir 1996-1997. 16 March 1999.
NRCD. 1982. North Carolina Clean Lakes Classification Study. North Carolina Department of Natural Resources and Community Development, Division of Environmental Management.
Preston, S.D., V.J. Bierman Jr., and S.E. Silliman. 1989. An evaluation of methods for the estimation of tributary mass loads. Water Resources Research, 25(6): 1379-1389.
Sauber, Jay (2001) Chlorophyll Meetings and Issues with Lab Data. Email correspondence with Trevor Clements, March, 1, 2001.
Smith, V. H. Prediction of Nuisance Blue-green Algal Growth in North Carolina Waters. WWRI Report No. 233. Water Resources Research Institute of the University of North Carolina,
Raleigh, N.C. 1987.
Tetra Tech. 1998. Randleman Lake Project: Support Documentation for Nutrient Load and Eutrophication Model. Prepared for Piedmont Triad Regional Water Authority. Tetra Tech, Inc.,
Research Triangle Park, NC.
Walker W.W., Jr. 1985. Empirical Methods for Predicting Eutrophication in Impoundments. Report 4—Phase II: Model Refinements. U.S. Army Corps of Engineers Technical Report E-B1-9.
Waterways Experiment Station, Vicksburg, MS.
Walker W.W., Jr. 1987. Empirical Methods for Predicting Eutrophication in Impoundments. Report 4—Phase III: Applications Manual. U.S. Army Corps of Engineers Technical Report E-B1-9.
Waterways Experiment Station, Vicksburg, MS.
Weiss, C. M., Francisco, D. E., and Campbell, P. H. Water Quality Study, B. Everett Jordan Lake, North Carolina Year III, December 1983-November 1984. Report to the Wilmington District
of the U.S. Army Corps of Engineers. 120 pp., October 1986.
Weiss, C. M., Francisco, D. E., and Campbell, P. H. Water Quality Study, B. Everett Jordan Lake, North Carolina Year II, December 1982-November 1983. Report to the Wilmington District
of the U.S. Army Corps of Engineers. 140 pp., July 1985.

Appendix A: Data Summary

Appendix B: Detailed Notes on Literature Review

Appendix C: Principal Components Analysis of Algal Data

Appendix D: FLUX Analyses of Tributary Loads

A printed version of Appendix D is not included due to its length. Data are available through the Triangle J Council of Governments.

Appendix E: BATHTUB Model Input Files

A printed version of Appendix E is not included due to its length. Data are available through the Triangle J Council of Governments.

August 2001 Jordan Lake Nutrient Response Modeling Project:
Existing Data Memorandum

Jordan Lake Nutrient Response Modeling Project: August 2001
Existing Data Memorandum

ii Tetra Tech

Tetra Tech iii

Tetra Tech i

Tetra Tech 55

Tetra Tech 60

Jordan Lake Nutrient Response Modeling Project: August 2001
Existing Data Memorandum

A-2 Tetra Tech

Tetra Tech Appendices

Jordan Lake Nutrient Response Modeling Project: August 2001
Existing Data Memorandum

Tetra Tech A-1

B-2 Tetra Tech

Tetra Tech B-1

C-2 Tetra Tech

Tetra Tech C-1

D-2 Tetra Tech

Tetra Tech D-1

E-2 Tetra Tech

Tetra Tech E-1

- 1989 Results -

- 1997 Results -

- 1990 Results -

N and P Data
Sheet1
Tot N Chart
Tot P Chart
Chart2
Trib ID
Trib Name
Mean Flow (hm3/yr)
Tot P (ppb)
Total N (ppb)
P (kg/y)
N (kg/y)
Haw River
haw river nps
lower new hope nps
middle new hope nps
upper new hope nps
Morgan Creek
New Hope Creek
Northeast Creek
Pittsboro WWTP
NPS
TOTAL
TotalP (kg/y)
Total N (kg/y)
Other
Mean Flow
(hm3/yr)
(ppb)
(kg/yr)
Total N
Total P
NUTRIENT LOADS FROM
FLUX ANALYSIS
OTHER NUTRIENT LOADS
(LU/LC Analysis)
1.00
1007.10
147.00
1138.00
148043.70
1146079.80
2.00
56.25
38.00
712.00
2137.50
40050.00
3.00
40.86
10.00
715.00
408.60
29214.90
4.00
38.50
39.00
709.00
1501.50
27296.50
5.00
74.60
40.00
696.00
2984.00
51921.60
6.00
34.72
138.00
2619.00
4791.36
90931.68
7.00
96.10
174.00
2187.00
16721.40
210170.70
8.00
24.08
332.00
3357.00
7994.56
80836.56
9.00
0.47
1712.00
10316.00
804.64
4848.52
2.00
56.25
38.00
712.00
2137.50
40050.00
3.00
40.86
10.00
715.00
408.60
29214.90
4.00
38.50
39.00
709.00
1501.50
27296.50
5.00
74.60
40.00
696.00
2984.00
51921.60
9.00
0.47
1712.00
10316.00
804.64
4848.52
7836.24
153331.52
148043.70
1146079.80
4791.36
90931.68
16721.40
210170.70
7994.56
80836.56
7836.24
153331.52
1007.10
147.00
1138.00
148043.70
1146079.80
34.72
138.00
2619.00
4791.36
90931.68
96.10
174.00
2187.00
16721.40
210170.70
24.08
332.00
3357.00
7994.56
80836.56
56.25
38.00
712.00
2137.50
40050.00
40.86
10.00
715.00
408.60
29214.90
38.50
39.00
709.00
1501.50
27296.50
74.60
40.00
696.00
2984.00
51921.60
0.47
1712.00
10316.00
804.64
4848.52
Haw River
Morgan Creek
New Hope Creek
Northeast Creek
Other
1146079.80
90931.68
210170.70
80836.56
153331.52
Jordan Lake Phosphorus Loads
Haw River
Morgan Creek
New Hope Creek
Northeast Creek
Other
148043.70
4791.36
16721.40
7994.56
7836.24
Haw River
Morgan Creek
New Hope Creek
Northeast Creek
Other
148043.70
4791.36
16721.40
7994.56
7836.24

N and P Data
Sheet1
Tot N Chart
Tot P Chart
Chart1
Trib ID
Trib Name
Mean Flow (hm3/yr)
Tot P (ppb)
Total N (ppb)
P (kg/y)
N (kg/y)
Haw River
haw river nps
lower new hope nps
middle new hope nps
upper new hope nps
Morgan Creek
New Hope Creek
Northeast Creek
Pittsboro WWTP
NPS
TOTAL
TotalP (kg/y)
Total N (kg/y)
Other
Mean Flow
(hm3/yr)
(ppb)
(kg/yr)
Total N
Total P
NUTRIENT LOADS FROM
FLUX ANALYSIS
OTHER NUTRIENT LOADS
(LU/LC Analysis)
1.00
1007.10
147.00
1138.00
148043.70
1146079.80
2.00
56.25
38.00
712.00
2137.50
40050.00
3.00
40.86
10.00
715.00
408.60
29214.90
4.00
38.50
39.00
709.00
1501.50
27296.50
5.00
74.60
40.00
696.00
2984.00
51921.60
6.00
34.72
138.00
2619.00
4791.36
90931.68
7.00
96.10
174.00
2187.00
16721.40
210170.70
8.00
24.08
332.00
3357.00
7994.56
80836.56
9.00
0.47
1712.00
10316.00
804.64
4848.52
2.00
56.25
38.00
712.00
2137.50
40050.00
3.00
40.86
10.00
715.00
408.60
29214.90
4.00
38.50
39.00
709.00
1501.50
27296.50
5.00
74.60
40.00
696.00
2984.00
51921.60
9.00
0.47
1712.00
10316.00
804.64
4848.52
7836.24
153331.52
148043.70
1146079.80
4791.36
90931.68
16721.40
210170.70
7994.56
80836.56
7836.24
153331.52
1007.10
147.00
1138.00
148043.70
1146079.80
34.72
138.00
2619.00
4791.36
90931.68
96.10
174.00
2187.00
16721.40
210170.70
24.08
332.00
3357.00
7994.56
80836.56
56.25
38.00
712.00
2137.50
40050.00
40.86
10.00
715.00
408.60
29214.90
38.50
39.00
709.00
1501.50
27296.50
74.60
40.00
696.00
2984.00
51921.60
0.47
1712.00
10316.00
804.64
4848.52
Haw River
Morgan Creek
New Hope Creek
Northeast Creek
Other
1146079.80
90931.68
210170.70
80836.56
153331.52
Jordan Lake Phosphorus Loads
Haw River
Morgan Creek
New Hope Creek
Northeast Creek
Other
148043.70
4791.36
16721.40
7994.56
7836.24
Haw River
Morgan Creek
New Hope Creek
Northeast Creek
Other
1146079.80
90931.68
210170.70
80836.56
153331.52