Rainfall Flood Study- 2018

Currituck County –
Coastal Resilience Rainfall
Flood Study
Currituck County, North Carolina
Contract #7284
Final Report
Date: August 14, 2018

Currituck County – Coastal Resilience Rainfall Flood Study | i
CONTRIBUTORS
Technical Leads:
Jason Giovannettone, Ph.D., P.E., Siva Sangameswaran, Ph.D., P.E., D.WRE.
Technical Contributions:
Chris Maderia
Project Manager:
Brian K. Batten, Ph.D.

This report was prepared by the County of Currituck under grant award # NA15NOS4190091 to
the Department of Environmental Quality, Division of Coastal Management from the Office for
Coastal Management, National Oceanic and Atmospheric Administration. The statements,
findings, conclusions, and recommendations are those of the author and do not necessarily
reflect the views of DEQ, OCM or NOAA.

Currituck County – Coastal Resilience Rainfall Flood Study | ii
REVISION HISTORY
April 23, 2018 – Draft Report
June 18, 2018 – Final Draft Report
June 29, 2018 – Final Draft Report, with copy edit
August 14, 2018 – Final Report (incorporating NDCEQ comments)

Currituck County – Coastal Resilience Rainfall Flood Study | iii
EXECUTIVE SUMMARY
This study was performed as part of the requirements of Contract No. 7284 between
Currituck County and the North Carolina Department of Environmental Quality (NCDEQ). A
flood is one of the most severe and potentially devastating natural disasters. Floods come in
many forms, such as river, coastal, and flash flooding. Whenever these types of floods occur,
long-term planning and adaptation, preparedness and response time are all critical factors in
reducing the overall impacts. Awareness of areas that are currently prone and will become
more prone to flooding in the future is essential to consider in short-term, as well as long-term
planning. A majority of planning activity related to resilience and climate adaptation, both in
the region and the State, has focused on coastal flooding and sea level rise. This study focuses
on inland flooding, which was identified as an area of limited research, and concerns all
communities. The study will also provide an example of work in which other communities can
engage. Such awareness comes from an understanding of a combination of not only regional
climatic factors, but also of non-climate factors that relate to natural, physical, and developed
characteristics.
The current study estimates flood susceptibility in Currituck County due to non-climatic
flood risk factors. Several quantitative and qualitative methods were considered to estimate
flood susceptibility. The final selected method involves performing a logistic regression. A
logistic regression is a statistical method for analyzing a dataset in which there are one or more
independent variables that determine a binary (yes or no) outcome. The method uses several
flood risk factors that could potentially affect the region and for which sufficient data were
available. Flood risk factors considered include elevation, slope, land curvature (concave,
convex, or flat), distance to water body, land cover, tree canopy density, surficial materials, soil
drainage class, and percent impervious surface. The objective was to link each of the flood risk
factors to the occurrence of flooding for a flood event having a recurrence interval of at least
100 years. This would be especially useful in identifying areas susceptible to flooding during
extreme events such as Hurricane Matthew, a hurricane that brought rainfall totals, flooding,
and a return period of up to 1 in 500 years to northern portions of the county (NOAA, 2017).
Due to recent extreme rainfall events over the region, high -resolution satellite images of spatial
flood inundation are not available. Alternatively it was decided to use the 100-year FEMA
Special Flood Hazard Area (SFHA) to develop a statistical model of coastal flood susceptibility
due to storm surge, which has historically been the leading cause of flooding in Currituck
County.

Currituck County – Coastal Resilience Rainfall Flood Study | iv
Prior to developing the flood susceptibility map for Currituck County, the entire county was
divided into more than 1,722,000 “30m x 30m” cells. Approximately 5 percent of the cells
(86,122 cells) covering a total area of near 30 square miles were randomly chosen throughout
Currituck County from which to extract data for each flood risk factor (refer to Table A-1 for
data source information). An equal number of these points were selected in locations that were
within and outside of the SFHA. The data for each flood risk factor were selected from all
locations using ArcGIS and associated with a ‘1’ if the location was within the floodplain and a
‘0’ otherwise. The resulting relationships between each factor and flood occurrence were
ingested into a logistic regression from which a regression coefficient was obtained for each
flood risk factor. The magnitude of the coefficients indicates the relative strength of each flood
risk factor’s influence on flooding. It was found that ‘elevation’, ‘land slope’, and ‘soil drainage
type’ have the most influence on flood susceptibility throughout the county. The final logistic
regression equation that was developed was then used to assign probabilities of flooding to all
locations to create an overall probability map of Currituck County. Probabilities were classified
within one of five classes: 0 – 20% (“very low risk”); 20 – 40% (“low risk”); 40 – 60%
(“medium risk”); 60 – 80% (“high risk”); and 80 – 100% (“very high risk”). Several types of
critical infrastructure were overlaid on the flood susceptibility map to identify those assets that
are most vulnerable to the 100-year flood. It was observed that several areas classified as “very
high risk” and “high risk” and that contained several types of critical infrastructure were
located in the central portion of the county as well as along the northern Outer Banks. Almost
all such high-risk areas in the central region fell within the FEMA SFHA, while several
locations exhibiting ‘medium’ to ‘very high’ risk along the Outer Banks were located outside the
SFHA; infrastructure within these ‘sub-region’ warrants additional attention regarding
potential flood mitigation efforts.
It should be noted that the FEMA 100-year SFHA is limited to the sub-watersheds of
greater than one square mile that FEMA chose to study with limited resources. Other limiting
factors are the age of the underlying studies illustrated by the FEMA maps (often more than
two decades old) and their focus on only areas where development existed or was imminently
anticipated. FEMA’s flood mapping is developed using physical models to perform hydrologic
and hydraulic analysis of a statistical flood event with a one percent chance of being equaled or
exceeded in any given year (referred to as the 100-year flood). The susceptibility maps from
this study provide a less expensive approach of covering all land area within the region. By
using the statistical modeling methodology described in this report it was possible to identify
the contribution of flood factors within the physically modeled FEMA SFHA and apply them to
the entire study region to identify areas thought to be vulnerable to flooding. One important
disclaimer is that the flood susceptibility map was created for present-day conditions and is

Currituck County – Coastal Resilience Rainfall Flood Study | v
only to be used for planning purposes; it is not intended to replace the FEMA mapping for
regulatory or flood insurance decisions.

Currituck County – Coastal Resilience Rainfall Flood Study | vi
TABLE OF CONTENTS

CONTRIBUTORS .............................................................................................................................i
REVISION HISTORY .................................................................................................................... ii
EXECUTIVE SUMMARY .............................................................................................................. iii
TABLE OF CONTENTS ................................................................................................................. vi
1. INTRODUCTION ....................................................................................................................... 1
2. LITERATURE REVIEW .............................................................................................................2
3. DATA AND METHOD ................................................................................................................4
3.1. Flood Risk Factors ......................................................................................................4
4. FLOOD INUNDATION ............................................................................................................ 17
5. LOGISTIC REGRESSION ........................................................................................................ 18
6. CRITICAL INFRASTRUCTURE .............................................................................................. 19
7. RESULTS ................................................................................................................................. 20
8. H & H STUDY PLANNING FRAMEWORK ............................................................................ 29
9. SUMMARY ............................................................................................................................... 31
10. FUTURE WORK ................................................................................................................ 32
11. REFERENCES ......................................................................................................................... 34
APPENDIX A: Input Data Metadata ............................................................................................ 37

Currituck County – Coastal Resilience Rainfall Flood Study | 1
1. INTRODUCTION
This study was performed as part of the requirements of Contract No. 7284 between
Currituck County and the North Carolina Department of Environmental Quality (NCDEQ). A
flood is one of the most severe and potentially devastating natural disasters. Flooding occurs in
many forms, such as river, coastal, and flash flooding. These flood events arise from a variety
of processes such as snow melt, severe precipitation events, storm surge, and on a more long-
term scale, relative sea-level rise. Whenever any of these types of flooding events occur, long-
term planning and adaptation, preparedness and response time are all critical factors in
reducing the overall impacts. Up until the year 2013, there had been no significant trend in the
severity of flooding in North Carolina (Peterson et al., 2013). However, flooding, according to
NOAA, was the second-most common natural hazard in North Carolina during the period 1996
- 2014 (first is thunderstorm & lightning), with a flood occurring an average of once every 7.6
days (USDOE, 2015). During the same period flooding resulted in the third largest annualized
property loss due to natural hazards behind hurricanes and tornados. Awareness of areas that
are more prone to flooding is essential to consider in long-term planning. It can also inform
short term strategies, such as the development of early warning mechanisms. Such awareness
comes from an understanding of a combination of not only climatic factors impacting the
region, but also of non-climate factors that relate to regional and site characteristics as well.
Various types of hydrological models can be used to model flood susceptibility and can be
categorized as conceptual, physically-based, or data-driven models. Conceptual models are
typically comprised of partial differential equations of continuity and momentum, which can
result in an accurate estimation of the internal mechanisms of the hydrological processes.
Conceptual models do require a large amount of calibration data and sophisticated analysis
tools, which is not within the scope of the current project. Physical models are based on an
understanding of complex physical processes, which can be effectively used for long-term flood
forecasting and reducing associated damages in a river basin. Rainfall/runoff models are one
of the most common models of this type. The disadvantage is that a large amount of time and
resources is often required to understand the complex physical processes that are a part of
these models. Physical models also require tremendous amounts of data for calibration and
validation along with long computation times.
Data-driven models alternatively use linguistic variables whose values include words or
phrases, rather than the conventional numerical variables used in the models described above.
Data-driven models extract information from the input/output datasets used to create the
model to develop a statistical correspondence between the data. Unlike physical models, they

Currituck County – Coastal Resilience Rainfall Flood Study | 2
do not require an understanding of the complex physical processes by which the data are
related, only an understanding of the hydrological and meteorological variables and regional
characteristics that influence flood risk. The assumption that future flooding occurrences will
occur under similar conditions as past events with respect to these non-climatic flood risk
factors needs to be made in such a model. The objective in most data-driven models is to
produce a list of relative weights for whatever flood risk factors have been identified. These
weights can then be used to produce a risk map. The method used to derive these weights
represents the major difference between the models. Therefore, there are two major decisions
that need to be addressed when using a data-driven model. The first decision involves the
specific method to be used, while the second decision identifies the flood risk factors that will
be addressed and weighted, if required, using the selected method.
2. LITERATURE REVIEW
Examples of data-driven models found in the literature include fuzzy logic (FL), artificial
neural networks (ANN), adaptive neuro-fuzzy interface system (ANFIS), and analytical
hierarchy process (AHP). The first of these models is Fuzzy Logic, which is set up using flood
risk factor membership functions and rules. The membership function for each factor
incorporates various classifications (e.g. high, medium, and low) of that factor. After the
variables are partitioned into their different “fuzzy” classes, an IF…THEN type of rule is set up
to establish the response of any combination of these “fuzzy” classes. For example, Gogoi and
Chetopa (2011) used a fuzzy rule-based model to forecast runoff in the Jiadhal Basin in
Northeast India. The authors used three flood risk factors (total monthly rainfall, mean
monthly temperature, previous month’s discharge) and three categories (e.g. high, medium,
and low) to describe projected runoff, resulting in a total number of 33 = 27 rules. Sets of
values for each variable were then tested against these rules to identify rules that are fulfilled to
a point that exceeds a certain threshold value. This final set of rules was then used to project
runoff based on values of the identified flood risk factors.
The second type of data-driven model is the Artificial Neural Network (ANN). ANNs
consist of layers of nodes or neurons, which include an input layer (number of neurons equals
the number of flood causative factors), an output layer (number of neurons equals the number
of types of desired outputs), and one or more hidden layers where algorithms are used to
model the complex relationships that exist between each flood causat ive factor and the
influence that they have on the output, which in the context of flooding would be water levels
and/or flow. Kia et al. (2012) used ANN to predict water levels and flood inundation using
seven potential flood causative factors: rainfall, slope, elevation, flow accumulation, soil, land

Currituck County – Coastal Resilience Rainfall Flood Study | 3
use/cover, and geology. Alternatively, the third model type is the Adaptive Neuro-Fuzzy
Inference System (ANFIS), which uses a combination of the numeric power of neural networks
and the verbal power of fuzzy logic. Such a model contains features of both types of models
such as learning and optimization abilities and “if-then” rule thinking to map an input space to
an output space. An example of this method was developed for the Barak River basin in
Northeast India by Ullah and Choudhury (2013). Issues with using an ANN, ANFIS, or any
method that incorporates neural networks, relates to their complexity and the high computing
power that is required to run the algorithms. Also, the quality of the resulting predictions in
many cases have been found to be inferior to other model types (Shortridge et al., 2016),
especially when the data used to validate the model contains values outside the range of data
used to train the model.
An alternative to the data-driven methods above is the Analytic Hierarchy Process (AHP).
AHP identifies potential flood risk factors and their associated weights using expert opinions
combined with geographical, statistical, and historical data. For example, Matori et al. (2014)
and Siddayao et al. (2014) used AHP in performing spatial assessments of floodplain risk in
northern Malaysia and the northern Philippines, respectively. Flood risk factors included
rainfall, geology, soil type, land use, population density, distance from riverbank and site
elevation and slope. The authors in both studies surveyed experts in the region and used the
survey results to develop weights for each factor. They combined the resulting weights with a
Geographical Information System (GIS) to produce a color-coded map representing various
levels of risk for each respective study region. The advantage of this method is that the final
product is a flood susceptibility map based on the combined experience of several years of
flooding events from various type of experts who are familiar with the region. The
disadvantage is that the results can be based on subjective and conflicting opinions, especially
when there are many flood risk factors being considered. This can be alleviated, however,
when using the overall factor weighting mechanisms that are typically used in AHP.
Other quantitative types of data-driven models include multivariate statistical analysis
(MSA) and multivariate logistic regression (MLR), or some combination of these. These
methods rely on numerical expressions that characterize the relationships between the
independent flood risk factors and flood inundation (Lee et al. 2012). The use of MSA typically
requires several strict assumptions to be made prior to the analysis and requires the
relationship between flooding and each flood risk factor to be considered independently from
any potential relationships between factors to develop weights for each factor. MLR can be
used to solve this issue by examining the relations between a dependent variable (e.g., whether
a location is flooded or not flooded) and any number of independent variables (e.g. flood risk

Currituck County – Coastal Resilience Rainfall Flood Study | 4
factors; Pradhan 2010). An advantage of MLR is that a separate analysis is not required to
estimate the weight of each flood risk factor as this functionality is already built into su ch
coding environments as R. Another advantage of MLR is that the variables can be continuous
and/or categorical and is straightforward to implement.
After considering the advantages and disadvantages of each modeling approach described
above and the project scope, MLR was selected for the current study.
3. DATA AND METHOD
The current work was conducted in Currituck County, NC, (location shown in Fig. 2-1), the
most northeastern county in North Carolina. Currituck County includes the northern
communities of North Carolina’s Outer Banks, which are separated from the mainland by
Currituck Sound. Currituck County is known for its pristine beaches, rich farmland, numerous
wildlife refuges, and the Corolla wild horses. In order to preserve these resources, which help
to maintain the county’s tourism industry, awareness of flooding susceptibility, particularly
due to storm surge and heavy precipitation as a result of hurricanes and other coastal storm
events and riverine flooding, is very important in this region. The potential impacts on tourism
can be economically devastating. Spatially, Currituck County consists of half-open water and
half land area. Therefore, the county takes its identity from its coastline, making it imperative
to plan properly to preserve and live in harmony with these resources.
3.1. Flood Risk Factors
There are several types of data that are required as independent variables when performing
any type of flood susceptibility study. These independent types of data represent parameters
that may contribute to flooding in a region, and are referred to as flood risk factors. Flood risk
factors utilized for flood susceptibility mapping should be measurable and collected throughout
the entire “Area of Influence” (AOI). However, they should not represent information that is
spatially uniform. Although there is no agreement on which risk factors are the standard for
any flood susceptibility analysis, there are factors that are more prominently used than others.
Some of the most common factors are listed in Table 2-1 along with the citations for a few of
the studies in which they were identified as influential.
A subset of the flood risk factors listed in Table 2-1 was chosen for the present study after
considering the availability, period of record, and completeness of each dataset as applied to
Currituck County. This subset includes the following factors: elevation, slope, land curvature,
land cover, distance to water body, tree canopy density, percent impervious surface, soil

Currituck County – Coastal Resilience Rainfall Flood Study | 5
drainage class, and geology. Abbreviations, sources, and the resolution/scale of each dataset
are given in Table A-1 in Appendix A. These flood risk factors were collected over the present
study’s AOI, which is defined as a polygon having boundaries that extend slightly outside of the
boundaries of Currituck County (refer to Fig. 2-1) and compiled into spatial databases using
the ArcGIS 10.2 software. All datasets were resized using linear interpolation to a 30 m x 30 m
grid comprised of a total of more than 1,722,000 cells.

Figure 2-1: Map showing the location of Currituck County (shaded red) and the “Area of Influence” (red dashed line) within the
northeast corner of the State of North Carolina.

Currituck County – Coastal Resilience Rainfall Flood Study | 6
Table 2-1: Flood risk factors and examples of studies where they have been considered.
Flood Risk Factors Literature
Temperature Gogoi & Chetia (2011)
Previous month’s discharge Gogoi & Chetia (2011)
Population Density Siddayao et al. (2014); Sinha et al. (2008); Zhang et al. (2005)
Distance from Riverbank Siddayao et al. (2014)
Landform:
slope/elevation/curvature
Matori et al. (2014); Siddayao et al. (2014); Tehrany et al. (2014); Lawal
et al. (2012); Saini & Kaushik (2012); Sinha et al. (2008); Zhang et al.
(2005)
Distance from access road Harrison & Qureshi (2003)
Land-use zoning Lawal et al. (2012); Harrison & Qureshi (2003)
Drainage density Lawal et al. (2012); Saini & Kaushik (2012)
Proximity to drainage Sinha et al. (2008)
Soil type/drainage Matori et al. (2014); Tehrany et al. (2014); Lawal et al. (2012); Saini &
Kaushik (2012); Yahaya (2008)
Distance from urban areas Harrison & Qureshi (2003)
Precipitation/rainfall Matori et al. (2014); Tehrany et al. (2014); Lawal et al. (2012); Gogoi &
Chetia (2011); Yahaya (2008); Zhang et al. (2005); Harrison & Qureshi
(2003)
Land cover/use & Vegetation Matori et al. (2014); Tehrany et al. (2014); Saini & Kaushik (2012);
Yahaya (2008)
Geology Matori et al. (2014); Tehrany et al. (2014)
Timber type/size/density Tehrany et al. (2014)

Prior to using each data set in the flood susceptibility analysis, each flood risk factor was
divided into classes. This is accomplished using the quantile method (Tehrany et al., 2014;
Umar et al., 2014; Papadopoulou-Vrynioti et al., 2013), which partitions each numerical data
set (e.g. elevation (0.0 – 59.8 m), land slope (0.0 – 13.5°), land curvature (-0.67 - 1.66), which
represents the shape of the land and identifies local low points (concave) and high points
(convex), tree canopy density (0.0 – 100.0%), distance to water body (0 – 25,000 m), and
percent impervious service (0.0 – 100.0%)) into classes containing the same number of
features or pixels. Partitioning the data in this manner ensures that data is included and that a
coefficient can later be determined for each flood risk factor class. For the purposes of this
study, each of these datasets was divided into 10 categories, excluding impervious surface,
which was divided into 6 classes, using the classifications given in Table 2-2; refer to Figs. 2-2
– 2-7, respectively, to view the spatial distributions of each class. Regarding the other datasets,
land cover was divided into 11 classes (Fig. 2-6); soil drainage class was divided into 8 classes
(Fig. 2-8); and surface geology was divided into 3 classes (Fig. 2-10).

Currituck County – Coastal Resilience Rainfall Flood Study | 7
Table 2-2: Regression coefficients for each class of each flood risk factor.
Factor Class Logistic
Coefficient
Factor Class Logistic
Coeffici
ent
a0 -- 12.6325 DIST(m) 0.00 – 0.00 0.0000
ELEV(m) < 0.94 0.0000 0.01 – 196.21 0.0000
0.94 – 1.41 0.6128 196.21 – 490.52 -0.4004
1.41 – 2.11 -0.3583 490.52 – 784.83 -0.9168
2.11 – 4.22 -3.0786 784.83 – 1,177.25 -0.9724
4.22 – 6.10 -6.4254 1,177.25 – 1,667.77 -1.1524
6.10 – 7.51 -9.3953 1,667.77 – 2,452.61 -1.4200
7.51 – 9.15 -11.3555 2,452.61 – 3,727.97 -1.5727
9.15 – 11.03 -12.0374 3,727.97 – 6,867.30 -1.2850
11.03 – 14.08 -11.1318 6,867.30 – 25,016.61 0.5664
14.08 – 59.82 -9.4493 IMP(%) 0.00 – 0.00 0.0000
CURV < -0.0247 0.0000 0.01 – 4.00 -1.0587
-0.0247 – -0.0156 0.4361 4.01 – 14.00 -0.7813
-0.0156 – -0.0064 0.7605 14.01 – 47.00 -0.4579
-0.0064 – 0.0027 0.8901 47.01 – 100.00 0.7882
0.0027 – 0.0118 1.0959 GEO Sediments - fine 0.0000
0.0118 – 0.0210 1.3195 Sediments - medium -0.2054
0.0210 – 0.0301 1.4729 Organic rich muck/peat 2.0944
0.0301 – 0.0392 1.6368 LAND developed, open space 0.0000
0.0392 – 0.0667 1.7871 dev., low intensity 0.6185
0.0667 – 1.656 1.8627 dev., medium/high intensity -0.0334
SLOPE 0.0000 – 0.0000 0.0000 barren land 1.0416
0.0001 – 0.0530 -2.6134 forest (ever./dec./mixed) -0.6885
0.0530 – 0.1060 -2.7826 shrub/scrub -0.8099
0.1060 – 0.1591 -2.6321 herbaceous -0.5439
0.1591 – 0.2121 -2.6346 hay/pasture -1.0040
0.2121– 0.3181 -2.6122 cultivated crops -1.1890
0.3181 – 0.4241 -2.4032 Woody wetlands -0.0645
0.4241 – 0.5832 -2.4183 Emergent herbaceous wetlands 0.7131
0.5832 – 0.9013 -2.5059 TREE(%) 0.00 – 0.00 0.0000
0.9013 – 13.5197 -2.3585 0.01 – 39.00 0.2313
SOIL unrated 0.0000 39.01 – 70.00 0.2404
very poorly drained -2.4643 70.01 – 83.00 0.2305
poorly drained -1.6092 83.01 – 89.00 0.0422
somewhat poorly -1.7349 89.01 – 92.00 0.0653
moderately well -1.6259 92.01 – 96.00 -0.0565
well drained -2.2206 96.01 – 98.00 -0.0014
somewhat
excessively
0.0000 98.01 – 99.00 -0.0862
excessively drained -0.6522 99.01 – 100.00 0.3439

Currituck County – Coastal Resilience Rainfall Flood Study | 8

Figure 2-2: Spatial distribution of flood risk factor “elevation”. (Source: https://earthexplorer.usgs.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 9

Figure 2-3: Spatial distribution of flood risk factor “curvature”. (Source: https://earthexplorer.usgs.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 10

Figure 2-4: Spatial distribution of flood risk factor “slope”. (Source: https://earthexplorer.usgs.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 11

Figure 2-5: Spatial distribution of flood risk factor “tree canopy density”. (Source: https://www.mrlc.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 12

Figure 2-6: Spatial distribution of flood risk factor “land cover”. (Source: https://www.mrlc.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 13

Figure 2-7: Spatial distribution of flood risk factor “distance to water”. (Source: https://www.mrlc.gov/)

Currituck County – Coastal Resilience Rainfall Flood Study | 14

Figure 2-8: Spatial distribution of flood risk factor “soil drainage”. (Source: https://sdmdataaccess.nrcs.usda.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 15

Figure 2-9: Spatial distribution of flood risk factor “impervious area”. (Source: https://www.mrlc.gov)

Currituck County – Coastal Resilience Rainfall Flood Study | 16

Figure 2-10: Spatial distribution of flood risk factor “surficial materials”. (Source: https://pubs.usgs.gov/ds/425/)

Currituck County – Coastal Resilience Rainfall Flood Study | 17
4. FLOOD INUNDATION
The overall objective is to develop relationships between flooding and all dependent
variables (flood risk factors). Therefore, a method is required to compare the values of each
factor at a point with whether flooding would be expected or not expected to occur at that point
for a specific flood event or flood recurrence frequency. This was initially going to be
accomplished using satellite images during a severe flood event that occurred within the last 15
years with sufficient spatial resolution to show maximum spatial inundation. Due to limited
access to high-quality satellite data during such an event, it was decided to focus on coastal
flood susceptibility and compare flood risk factors to flood inundation as defined by the 100-yr
FEMA Special Flood Hazard Area (SFHA) for the region (Fig. 2-11). It was assumed that
estimated correlations between flood risk factors using the SFHA and flood inundation due to a
coastal event are similar regardless of whether a specific observed flood event is used as
compared to the SFHA. Alternatively, if flood susceptibility due to pluvial flooding was
desired, observations from actual events would be required. Flood inundation data from the
SFHA were compiled into a spatial database using the ArcGIS 10.2 software and resize d to a 30
m x 30 m grid; the grid of Currituck County was constructed using the Area of Influence (AOI)
shown in Fig. 2-1.

Figure 2-11: 100-year FEMA national flood hazard layer over Currituck County.

Currituck County – Coastal Resilience Rainfall Flood Study | 18
5. LOGISTIC REGRESSION
Logistic regression, which is a statistical method for analyzing a dataset in which there are
one or more independent variables that determine a binary (yes or no) outcome, is then
implemented to develop a specific formula that measures the probability of flood inundation
throughout the Area of Interest (AOI) during the 100-year flood event. This is accomplished by
designating several points throughout the AOI as testing points from which the logistic
regression will be derived. Approximately 5 percent of the total number of “30m x 30m” cells
that make up Currituck County in this analysis, or 86,122 cells equal to an area of near 30
square miles, were randomly chosen throughout Currituck County with the stipulation that an
equal number of those points (43,063) were within and outside of the SFHA (illustrated in Fig.
2-12).
Flood data for all points consisted of either a 0 or a 1 to represent whether a location was
not flooded or flooded, respectively; these values represented the dependent variable (L) in the
logistic regression:
𝑙𝑛(𝑝
1−𝑝)=𝐿= 𝑎0 +𝑎1 𝑥1 +𝑎2 𝑥2 +⋯+𝑎𝑛𝑥𝑛, (1)
where p is the probability of flooding. All flood risk factor data at each location was categorized
into classes according to the class ranges designated in Table 2-2 and represented the
independent variables (x1 to xn; n = 9) in the logistic regression in Eq. 1. In some cases, the
‘land cover’, ‘soil class’, and/or ‘surficial materials’ risk factors were classified as ‘open water’
and/or the ‘distance to water’ was equal to zero even though the location was located outside of
any body of water. These points were eliminated from the analysis, which resulted in the total
number of points being utilized in the study to be 85,681. The independent and dependent
variables were then analyzed using the logistic regression function in R Statistical Software (R)
to determine the regression intercept (a0) and the coefficients (a1 to an; n = 9) for each flood
risk factor in Eq. 1.
After the coefficients of the logistic regressions are determined for each class of each flood
risk factor, the following equation, which is derived from Eq. 1, is used to calculate the
probability of flooding at each map grid cell in the final flood susceptibility map. It should be
noted that all flood risk factors are used but that for each flood risk factor only one coefficient
is used that corresponds to the appropriate factor class (see Table 2 -2) at each map grid cell:

Currituck County – Coastal Resilience Rainfall Flood Study | 19
𝑝= 𝑒𝐿
(1 +𝑒𝐿)⁄ . (2)

Figure 2-12: Map of a zoomed-in portion of Currituck County showing the distribution of sampling points used to train the
logistic model in relation to the boundary of the SFHA. Green points represent locations where flooding did not occur, while
red points represent locations where flooding did occur.
6. CRITICAL INFRASTRUCTURE
The final step of the methodology related to the development of the flood susceptibility map
involves identifying vulnerable critical infrastructure. The geographic information (GIS) data
sets include:
• Dams
• Airports
• Hospitals and other health-related facilities
• Fire and police stations
• County facilities
• Private and public K – 12 schools

Currituck County – Coastal Resilience Rainfall Flood Study | 20
• Major routes
• Bridges
• Railroads

Data sets and sources related to critical infrastructure throughout Currituck County and
that were used in the current study are given in Table A-2 in Appendix A. All critical
infrastructure datasets were clipped to the regional boundaries of Currituck County and
overlaid onto the final flood susceptibility map.
7. RESULTS
The coefficients resulting from the logistic regression are given in Table 2-2 for each class of
each flood risk factor. The greater the magnitude of the coefficient, the stronger the impact of
that risk factor class on flooding in the AOI. The average regression coefficient values for all
flood risk factors are illustrated in Figure 3-1. There are three flood risk factors that stand out
as having a dominant correlation with flood susceptibility throughout Currituck County:
‘elevation’ (ELEV), ‘land slope’ (SLOPE), and ‘soil drainage class’ (SOIL). The fact that
elevation appears to have the highest influence is not surprising due to the impacts of storm
surge within the region. The results of the logistic regression for the initial set of data points
were then applied to all map grid cells in Currituck County to produce a flood susceptibility
map for the entire region applicable to the 100-year flood event (Fig. 3-2). Flood susceptibility
values are plotted as the percent chance that each 30 m x 30 m grid cell will be flooded and
then classified into five categories according to the color scale shown in the figure: very low
risk (0 – 20%), low risk (20 – 40%), medium risk (40 – 60%), high risk (60 – 80%), and very
high risk (80 – 100%). The largest areas of ‘very high’ and ‘high’ susceptibility are located on
Knott’s Island, the Outer Banks, and the mainland along Currituck Sound, as well as along the
Northwest and North Rivers and their tributaries. There is also an area of high susceptibility in
the northwest corner of the region due to a number of flood risk factors not related to
proximity to a water body, such as land cover, soil drainage, and high tree canopy density. The
fact that this area consists of high-density woody wetlands (Figs. 2-5 and 2-6) and is
characterized by very poorly drained soil conditions (Fig. 2-8) contributes to its high flood
susceptibility. It is important to note that at locations where data for one or more flood risk
factors was unavailable, we were unable to calculate the flood susceptibility, which was
assigned as "undetermined".

Currituck County – Coastal Resilience Rainfall Flood Study | 21

Figure 3-1: Average absolute value of the logistic regression coefficients computed for each flood risk factor.

Currituck County – Coastal Resilience Rainfall Flood Study | 22

Figure 3-2: Flood susceptibility map of Currituck County using the 100-year FEMA SFHA. Levels represent probabilities of
flooding: Very Low: 0 – 20%; Low: 20 – 40%; Medium: 40 – 60%; High: 60 – 80%; Very High: 80 – 100%.

Currituck County – Coastal Resilience Rainfall Flood Study | 23
When comparing the susceptibility mapping to the FEMA SFHA, it is important to
understand key distinctions between the two. The SFHA is limited to the sub-watersheds of
greater than one square mile that FEMA chose to study with limited financial and
computational resources. Other limiting factors are the age of the underlying studies
illustrated by the FEMA maps (often more than two decades old) and their focus on only areas
where development existed or was imminently anticipated. FEMA’s flood mapping is
developed using physical models to perform hydrologic and hydraulic analysis of a statistical
rainfall event with a one percent chance of being equaled or exceeded in any given year
(referred to as the 100-year flood). In general terms, hydrologic analysis is the study of
transforming rainfall amount into quantity of runoff. Hydraulic analysis takes that quantity of
water and uses a physical model to route it through existing terrain, while considering such
factors as topography and vegetative density. This modeling is referred to as “detailed
analysis.” Some areas are studied by “approximate methods.” In general, areas studied by
approximate methods use a simplified hydrologic analysis methodology and route runoff
quantity through best available topography alone.
The susceptibility maps from this study provide a less expensive method of covering all land
area within the region. By using the statistical modeling methodology described in this report it
was possible to identify the contribution of flood factors within the physically modeled FEMA
SFHA and apply them to the entire study region to identify areas thought to be vulnerable to
flooding. One important disclaimer about the flood susceptibility map is that it was created for
present-day conditions and is only to be used for planning purposes. It is not intended to
replace the FEMA mapping for regulatory or flood insurance decisions.
The scale of the flood susceptibility map and data are most appropriately used at the
regional scale. However, use of the data at the municipal scale should allow local officials to
examine areas of concern for planning purposes. As more accurate input datasets (e.g. higher
resolution LiDAR data and imagery) become available, they can be easily incorporated into an
updated flood susceptibility analysis. Higher resolution input datasets also allow smaller areas
to be analyzed in more detail if desired (e.g. the City of Coinjock and other communities along
Highway 158 in the center of the map, which are dominated by areas of ‘very high’ flood
susceptibility in Fig. 3-2).
Data sets for various types of critical infrastructure (listed in Table A-2) were obtained and
overlaid onto the final flood susceptibility map for Currituck County (Fig. 3-3). Several critical
infrastructure, and a large portion of the major routes and railroad in the central portion of the
county, are included within the ‘high’ and ‘very high’ risk areas of 100-year flood susceptibility.

Currituck County – Coastal Resilience Rainfall Flood Study | 24
It is also observed that almost all areas identified as having “high” to “very high” flood
susceptibility to the 100-year flood are included in the FEMA SFHA. There are a few
exceptions, particularly along the northern Outer Banks as can be observed using the zoomed-
in maps shown in Fig. 3-4 (locations within the county are indicated by the rectangles in Fig. 3-
3). Figures 3-4a – c reveal several areas where flood susceptibility ranges from ‘medium’ to
‘very high’ that are located outside of the SFHA moving from north (Fig. 3-4a) to south (Fig. 3-
4c). In particular, almost half of the area located outside of the SFHA in Fig. 3 -4a is classified
as being susceptible to 100-year flood events. In addition to road, there are critical
infrastructure located in areas of high susceptibility (Figs. 3-4b and c) for which additional
flood mitigation efforts may be warranted.
It should be noted that as the flood susceptibility map provides a guide for future planning
and flood preparedness related to critical infrastructure and other facilities, it is not meant to
provide additional information to be used for flood insurance or regulatory purposes; this is
the purpose of the FEMA map or hatched area in Fig. 3-3.

Currituck County – Coastal Resilience Rainfall Flood Study | 25

Figure 3-3: Locations of various vulnerable critical infrastructure relative to areas of ‘Medium’ (dark green), ‘High’ (dark red),
and ‘Very High’ (red) flood susceptibility. The 100-year FEMA SFHA (hatched) is also included for reference and comparison.
Boxes represented sub-regions analyzed in more detail in Fig. 3-4.

Currituck County – Coastal Resilience Rainfall Flood Study | 26

(a)

Currituck County – Coastal Resilience Rainfall Flood Study | 27

(b)

Currituck County – Coastal Resilience Rainfall Flood Study | 28

(c)
Figure 3-4: Comparison of locations having ‘medium’ (yellow), ‘high’ (orange), and ‘very high’ (red) flood susceptibilities that lie
outside of the FEMA SFHA (red-hatched area) for three sub-regions (a, b, and c) along the northern Outer Banks; critical
infrastructure is overlaid on the map. Specific locations of each sub-region are shown by the boxes in Fig. 3-3.

Currituck County – Coastal Resilience Rainfall Flood Study | 29
8. H & H STUDY PLANNING FRAMEWORK
The following analytic framework was developed to assist the task of prioritization and
planning for detailed hydrological and hydraulic (H & H) studies. Figure 3-5 outlines the
process for identifying drainage basins and performing detailed H & H analysis aimed at
developing hazard mitigation strategies and solutions.

Figure 3-5: Detailed H & H Analysis Process Flow

Currituck County – Coastal Resilience Rainfall Flood Study | 30
The following is a set of scoping questions incorporating the priorities, concerns and needs
of various agencies and stakeholders, to be used as evaluation criteria for detailed H & H study
area selection:
1. What are the main areas of susceptibility identified based on watershed conditioning
factors (results of this study)?
2. What public facilities are in areas of substantial vulnerability and are currently
unprotected / under protected?
3. What critical infrastructure facilities (hospitals, emergency response facilities including
fire stations, shelters, etc.) are in areas of substantial vulnerability, and are currently
unprotected / under protected.
4. Roadway infrastructure criteria:
a. How many, and what are the locations of the State and County owned roadways
that connect rural areas of the County to interior parts of the County, which are
highly vulnerable?
b. What are the existing drainage issues in these roadway networks (undersized
culverts, un/under-maintained roadway segments, bridges and culverts, which
form the hydraulic pathway of least resistance resulting in flood susceptibility?
5. What are the locations of County Waste Water Treatment Plants (WWTPs), water
distribution facilities and systems, and landfill (sanitary and hazardous waste) sites that
are in the areas of susceptibility?
6. Are there known issues of sewer and stormwater mixing resulting in hazardous
conditions for the environment? If so, what are the sources of and approximate (if
known) extent of this type of impact stormwater (only) inundation areas?
7. What are the social assets (historic structures, religious buildings, other health care
facilities, etc.) that are identified within the areas of rainfall susceptibility?
8. What are the areas of combined coastal, fluvial (riverine), and pluvial (rainfall)
influence?
9. What are the high priority assets and resources from an environmental, ecosystem and
habitat stand point (including wetlands and shorelines)? Based on inputs received from
The Nature Conservancy’s Office of Coastal Engagement, the following are some items
to consider:
a. Areas in the Northwest part of the County and north river watershed.
b. Areas identified as high priority under NC Heritage significant natural areas.
c. Areas identified as “resilient”, i.e. ranked as equal to greater than average in The
Nature Conservancy’s “Resilient Coastal Sites”
d. Areas of migration space / Sea Level Rise (SLR) buffer.

Currituck County – Coastal Resilience Rainfall Flood Study | 31
10. What are the current, near-term and long term development goals and / or restrictions
by the County, including full service areas?
11. What are the possible ways to obtain and /or collect missing data that resulted in
“undetermined” flood susceptibility in this study?
12. What are the key climate variability aspects (sea level rise, increased precipitation,
extreme temperatures) to be considered while identifying and prioritizing basins for
detailed watershed studies?
13. How does the County intend to best use the findings of this study to articulate, educate
and invite participation from the residents and agencies involved to plan for and
implement long term resiliency measures?
9. SUMMARY
A flood is one of the most severe and potentially devastating natural disasters. Awareness
of areas that are currently prone and will become more prone to flooding in the future is
essential to consider in short-term, as well as long-term, planning. Such awareness comes
from an understanding of a combination of not only regional climatic factors, but also of non-
climate factors that relate to regional and site characteristics.
The current study estimated flood susceptibility within Currituck County, North Carolina,
due to non-climatic factors. The method used to look at flood susceptibility involved
performing a logistic regression to determine the relationship between several flood risk
factors and flooding at the 100-year recurrence level within the county. It was found that
‘elevation’, ‘land slope, and ‘soil drainage class’ have the most influence on flood susceptibility
in the region. The coefficients that resulted from the logistic regression were then used to
create an overall flood susceptibility map for Currituck County onto which various types of
critical infrastructure were overlaid. Large areas with “very high” and “high” susceptibility
where such infrastructure are located were identified within the central portion of the county
and well as the northern Outer Banks. Although the regional data is not at a scale large enough
for local determinations, these hotspot areas warrant further consideration for future localized
flood susceptibility mapping if future suitable data sets become available and further
consideration at the municipal resiliency planning level.
Minimal differences were observed between the 100-year susceptibility map and the 100-
year FEMA SFHA throughout most of the county because land characteristics that correlated
with flooding did not extend beyond the limits of the SFHA. With that said, there were a few
areas of ‘medium’ to ‘very high’ susceptibility found along the Outer Banks where land
characteristics similar to those found within the SFHA extended outside of the limits of the

Currituck County – Coastal Resilience Rainfall Flood Study | 32
SFHA as well and that warrant additional attention. One important disclaimer about the flood
susceptibility map is that it was created for present-day conditions and is only to be used for
planning purposes. There are several prominent factors that could affect the future flood
susceptibility map: changes in impervious area (through urbanization), a higher sea level (for
coastal areas) and heavier precipitation. A future flood susceptibility map can be created by
studying how these factors are expected to change. However, it is expected that the present -day
flood susceptibility map provides an excellent relative foundation from which to consider
future changes. In other words, it is logical to assume that higher-risk present-day regions will
remain as higher-risk regions in the future.
These conclusions demonstrate the importance of determining which present-day
recurrence intervals (e.g. 100-year) are important for land use and recovery planning, hazard
mitigation, zoning, design standards and/or flood warning plans. Socioeconomic models can
then be built to show how a more frequent occurrence of such events will impact response
and/or recovery costs.
10. FUTURE WORK
Projects and studies that utilize novel methods in accomplishing their final objectives
typically identify several additional new directions in which to extend the work as well as
additional questions that come up because of the analysis and conclusions. The current
project is no exception with the following list providing potential avenues for future work:
• Utilize local expert and resident experiences related to flooding in the region to ground-
truth the 100-year flood susceptibility map that was developed in the current study.
• Maintain awareness of data collection for future events. Given the increase in forecast
skill of severe floods, it may be possible for Currituck County to work with its
neighbors/partners to make sure that any future flood inundation events are well
sampled by specialized satellite and/or synthetic aperture radar missions. These would
provide the horizontal resolution to significantly enhance the current model past the 30-
m grid size.
• Create additional flood susceptibility maps for more frequent flood exceedance
frequencies using the method used for the 100-year flood events. This is limited by the
availability of satellite data during maximum inundation caused by the flood, but images
for very frequent events (e.g. 5-year) should be available and would provide inundation
information for floods that are considered a frequent annoyance rather than a
potentially rare disaster.
• Re-run the analysis for future flood events. If and when a flood event occurs in the
future within Currituck County and resources and satellite imagery permitting, recreate

Currituck County – Coastal Resilience Rainfall Flood Study | 33
a flood susceptibility map for the exceedance frequency associated with the event. The
final goal would be to analyze a sufficient number of events of varying frequencies to
enable interpolation of the risk factor regression coefficients for any flood event
exceedance frequency.
• Identify, obtain and collect (if necessary) missing data that resulted in the assignment of
“undetermined” flood susceptibility.
• Encourage the development of improved datasets related to flood risk factors that were
identified as having substantial impacts on flooding in each sub-region; this would
include the flood-risk factors ‘elevation’, ‘distance to water’, and ‘land cover’. Improved
resolutions (e.g. 30 meters to 1 meter) of each input dataset would contribute
substantially to improved flood susceptibility maps at any desired exceedance
frequency.
• As resources permit, flood susceptibility map(s) should be revised, which includes
rerunning the analysis described in this report, as improved datasets of flood risk factors
become available.
• Using the framework developed as part of this study, identify areas for detailed H & H
analysis based on priority ranking, inter-agency and multi-stakeholder collaboration.

Currituck County – Coastal Resilience Rainfall Flood Study | 34
11. REFERENCES
Brekke, L., B. L. Thrasher, E.P. Maurer, and T. Pruitt. (2013) "Downscaled CMIP3 and CMIP5
climate projections: release of downscaled CMIP5 climate projections, comparison with
preceding information, and summary of user needs." US Department of the Interior, Bureau
of Reclamation, Technical Service Center, Denver, Colorado, USA.
Castro, C.L., H.I. Chang, L.O. Mearns, and M.S. Bukovsky. (2015) “Trend of climate extremes
in North America: A comparison between dynamically downscaled CMIP3 and CMIP5
simulations.” In American Geophysical Union Fall Meeting Abstracts.
Gogoi, S., and B.C. Chetia. (2011) “Fuzzy Rule-based Flood Forecasting Model of Jiadhal River
Basin, Dhemaji, Assam, India,” International Journal of Fuzzy Mathematical Systems, 1(1),
59 – 71.
Flato, G., J. and co-authors. (2013) Evaluation of Climate Models. In: Climate Change 2013:
The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report
of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
Harrison, S., and M. Qureshi. (2003) “Application of the analytic hierarchy process to riparian
revegetation policy options,” Smallscale Forest Economics, Management and Policy, 2(3),
441 – 458.
Kia, M.B., S. Pirasteh, B. Pradhan, A.R. Mahmud, W.N.A. Sulaiman, and A. Moradi. (2012) “An
artificial neural network model for flood simulation using GIS: Johor River Basin, Malaysia,”
Environmental Earth Sciences, 67, 251–264.
Kam, J., and Co-authors. (2016) Multimodel Assessment of Anthropogenic Influence on
Record Global and Regional Warmth During 2015 [in “Explaining Extremes of 2015 from a
Climate Perspective”]. Bulletin of the American Meteorological Society, 97 (12), S4 –S8,
doi:10.1175/BAMS-D-16-0149.
Lawal, D., A. Matori, A. Hashim, K. Yusof, and I. Chandio. (2012) “Detecting flood susceptible
areas using GIS-based analytic hierarchy process,” Proceedings of the International
Conference on Future Environment and Energy, 28, 3 – 4.
Lee, M. J., J. E. Kang, and S. Jeon. (2012) “Application of frequency ratio model and validation
for predictive flooded area susceptibility mapping using GIS.” In: Geoscience and Remote
Sensing Symposium (IGARSS), 2012 IEEE International. Munich, 895–898.
Matori, A.N., D.U. Lawal, K.W. Yusof, M.A. Hashim, and A.-L. Balogun. (2014) “Spatial
Analytic Hierarchy Process model for flood forecasting: An integrated approach,” IOP
Conference Series: Earth and Environmental Science, 20, 1 – 7.
Melillo, J.M., T. Richmond, and G.W. Yohe (eds.). (2014) “Climate Change Impacts in the
United States: The Third National Climate Assessment.” U.S. Global Change Research
Program.
Menne, M.J., et al. (2012) “An overview of the global historical climatology network -daily
database.” Journal of Atmospheric and Oceanic Technology, 29, 897-910.

Currituck County – Coastal Resilience Rainfall Flood Study | 35
NOAA (2017) “Service Assessment: October 2016 Hurricane Matthew,” 59 p.
Perica, S., et al. (2015) “NOAA Atlas 14 Volume 10 Version 2, Precipitation-Frequency Atlas of
the United States, Northeastern States.” NOAA, National Weather Service, Silver Spring,
MD.
Peterson, T.C., et al. (2013) “Monitoring and understanding changes in heat waves, cold
waves, floods and droughts in the United States: State of knowledge.” Bulletin of the
American Meteorological Society, 94, 821-834.
Pradhan, B., and S. Lee. (2010a) “Delineation of landslide hazard areas on Penang Island,
Malaysia, by using frequency ratio, logistic regression, and artificial neural network models,”
Environ. Earth. Sci., 60, 1037–1054.
Prein, A.F., et al. (2017) “The future intensification of hourly precipitation extremes.” Nature
Climate Change, 7, 48-52.
Saini, S.S., and S.P. Kaushik. (2012) “Risk and vulnerability assessment of flood hazard in part
of Ghaggar basin: A case study of Guhla block, Kaithal, Haryana, India,” International
Journal of Geomatics and Geosciences, 3(1), 10.
Shortridge, J.E., S.D. Guikema, and B.F. Zaitchik. (2016) “Machine learning methods for
empirical streamflow simulation: a comparison of model accuracy, interpretability, and
uncertainty in seasonal watersheds,” Hydrology and Earth System Sciences, 20, 2611 –
2628.
Siddayao, G.P., S.E. Valdez, and P.L. Fernandez. (2014) “Analytic Hierarchy Process (AHP) in
Spatial Modeling for Floodplain Risk Assessment,” International Journal of Machine
Learning and Computing, 4(5), 450 – 457.
Sinha, R., G.V. Bapalu, L.K. Singh, and B. Rath. (2008) “Flood risk analysis in the Kosi River
Basin, North Bihar using multi-parametric approach of analytical hierarchy process (AHP),”
Journal of The Indian Society of Remote Sensing, 36, 293 – 307.
Tehrany, M.S., M.-J. Lee, B. Pradhan, M.N. Jebur, and S. Lee. (2014) “Flood susceptibility
mapping using integrated bivariate and multivariate statistical models,” Environmental
Earth Sciences, 72(10), 1 – 15.
Themeßl, M.J., Gobiet, A., and A. Leuprecht. (2011) “Empirical‐statistical downscaling and
error correction of daily precipitation from regional climate models.” International Journal
of Climatology, 31, 1530-1544.
Ullah, N., and P. Choudhury. (2013) “Flood Flow Modeling in a River System Using Adaptive
Neuro-Fuzzy Inference System,” Environmental Management and Sustainable
Development, 2(2), 54 – 68.
United States Department of Energy (USDOE). (2015) “State of North Carolina Energy Sector
Risk Profile,” 8 pp.
Yahaya, S. (2008) “Multicriteria analysis for flood vulnerable areas in Hadejia-Jama’are River
Basin, Nigeria,” American Society for Photogrammetry and Remote Sensing, 6.
Zhang, H., J. Zhang, and J. Han. (2005) “The assessment and regionalization of
flood/waterlogging disaster risk in middle and lower reaches of Liao River of Northeast

Currituck County – Coastal Resilience Rainfall Flood Study | 36
China,” Proceedings of the Fifth Annual IIASA-DPRI Forum on Integrated Disaster Risk
Management, 103 – 118.

Currituck County – Coastal Resilience Rainfall Flood Study | 37
APPENDIX A: Input Data Metadata
Table A-1: Flood risk factors and flood event data with data source and resolution/scale.
Flood Risk Factors Source (year) Resolution/Scale URL for Data Access
Land Cover
(LAND)
USGS (2011) 30 meters https://www.mrlc.gov/
Elevation (ELEV);
Slope (SLOPE);
Curvature (CURV)
USGS (2014;
2011)
30 meters https://earthexplorer.usgs.gov/
Distance from
Water (DIST)
USGS (2014) 1:24,000 https://www.mrlc.gov/
Soil Drainage
(SOIL)
USDA-NRCS
(current)
varies https://sdmdataaccess.nrcs.usda.gov/
Vegetation density
(VEG)
USGS (2011) 30 meters https://www.mrlc.gov/
Impervious Surface
(IMP)
USGS (2011) 30 meters https://www.mrlc.gov/
Surface Geology
(GEO)
USGS (2009) 1:24,000 https://pubs.usgs.gov/ds/425/
FEMA 100-year
Hazard Area
DHS/FEMA
(2016)
1:12,000 https://catalog.data.gov/dataset/national-
flood-hazard-layer-nfhl

Table A-2: Critical infrastructure data sets used in the current study with data source and link.
Infrastructure Source (Year) URL for Data Access
Airports Currituck County
(2018)
http://co.currituck.nc.us/geographic-information-
services/
Bridges NCDOT (2018) https://connect.ncdot.gov/resources/gis/pages/gis-data-
layers.aspx
County
Facilities
Currituck County
(2018)
http://co.currituck.nc.us/geographic-information-
services/
Fire Stations Currituck County
(2018)
http://co.currituck.nc.us/geographic-information-
services/
Health NC OneMap
(2018)
http://data.nconemap.com/geoportal/catalog/main/hom
e.page
Police Stations NC OneMap
(2017)
http://data.nconemap.com/geoportal/catalog/main/hom
e.page
Railroads NCDOT (2018) https://connect.ncdot.gov/resources/gis/pages/gis-data-
layers.aspx
Major Routes NCDOT (2018) https://connect.ncdot.gov/resources/gis/pages/gis-data-
layers.aspx
Schools Currituck County
(2018)
http://co.currituck.nc.us/geographic-information-
services/

Analysis of Historical and Future Heavy Precipitation | 38