Assessing Indicators of Community Participation, A Dasymetric and

Loading...

University of South Carolina

Scholar Commons Theses and Dissertations

5-2017

The Community Rating System: Assessing Indicators of Community Participation, A Dasymetric and Sovi Approach Zachary P. Landis University of South Carolina

Follow this and additional works at: http://scholarcommons.sc.edu/etd Part of the Arts and Humanities Commons, and the Geography Commons Recommended Citation Landis, Z. P.(2017). The Community Rating System: Assessing Indicators of Community Participation, A Dasymetric and Sovi Approach. (Master's thesis). Retrieved from http://scholarcommons.sc.edu/etd/4094

This Open Access Thesis is brought to you for free and open access by Scholar Commons. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact [email protected]

THE COMMUNITY RATING SYSTEM: ASSESSING INDICATORS OF COMMUNITY PARTICIPATION, A DASYMETRIC AND SOVI® APPROACH by Zachary P. Landis Bachelor of Science United States Military Academy, 2008

Submitted in Partial Fulfillment of the Requirements For the Degree of Master of Arts in Geography College of Arts and Sciences University of South Carolina 2017 Accepted by: Kirstin Dow, Director of Thesis Dwayne Porter, Reader Susan L. Cutter, Reader Cheryl L. Addy, Vice Provost and Dean of the Graduate School

© Copyright by Zachary P. Landis, 2017 All Rights Reserved

ii

ABSTRACT The National Flood Insurance Program (NFIP) was established to provide affordable insurance to property owners and encourage communities to adopt and enforce floodplain management, primarily through the Community Rating System (CRS). The CRS awards points to communities for adopting a variety of activities in support of floodplain management. One area of interest in the CRS program is understanding differences in the types of communities participating and the degree of their participation. Research on the NFIP’s CRS tends to focus on community program participation in reducing flood losses and indicators of participation. Much of this research was performed prior to 2013 revisions to the CRS points system and considers the characteristics of the full CRS community, rather than just of the floodplain occupants, and uses single census factors. This study considers how robust these findings are given updates to the CRS points system and alternative methodological approaches. This research asks three main questions. Do previously identified indicators of community CRS participation remain useful, for overall points and for points within each CRS series? Are there significant differences between the Social Vulnerability Index (SoVI®) (Cutter et al. 2003) and the individual factors (educational attainment, housing value, population density) in the correlation to the CRS points? How does the application of a dasymetric approach to identify populations and their characteristics within the 100year floodplain compare to the correlations of SoVI® and individual factors (educational attainment, housing value, population density) with community CRS participation

iii

calculated at the community level? The analysis of indicator verification and potential was conducted using Pearson’s r and compared using Fisher’s z transformation. The comparison between the whole community and floodplain in the community was done through a paired t-test. The results confirmed the strength of housing value and education attainment as strong indicators of CRS participation post-2013 revisions, but population density was not found to be significant. The SoVI® was found to have indicator strength comparable to both housing value and educational attainment for indicating CRS participation. The SoVI® finding indicated that the more vulnerable communities tend to have lower levels of CRS participation. The results indicated that a dasymetric approach has limited value in examining the CRS within this study region.

iv

TABLE OF CONTENTS ABSTRACT .......................................................................................................................... iii LIST OF TABLES .................................................................................................................. vi LIST OF FIGURES ................................................................................................................ vii CHAPTER 1 INTRODUCTION ...................................................................................................1 CHAPTER 2 LITERATURE REVIEW .........................................................................................9 CHAPTER 3 METHODS .........................................................................................................17 CHAPTER 4 RESULTS ...........................................................................................................24 CHAPTER 5 DISCUSSION ......................................................................................................30 CHAPTER 6 CONCLUSION ....................................................................................................33 REFERENCES .......................................................................................................................36 APPENDIX A: LIST OF CRS COMMUNITIES STUDIED ...........................................................40 APPENDIX B: DATA USED ...................................................................................................43

v

LIST OF TABLES Table 1.1 2013 to Present: Study Region CRS Points by Series and Activity ....................2 Table 1.2 2013 to Present: National CRS Points by Series and Activity ............................3 Table 1.3 CRS Classes ........................................................................................................6 Table 2.1 Summary of Related Research ...........................................................................11 Table 2.2 Studies Showing Census Factor Correlation to Aspects of the CRS .................12 Table 4.1 Correlation Values .............................................................................................25 Table 4.2 Statistically Significant r Values .......................................................................25 Table 4.3 Paired T-Test between Floodplain and Whole Community ..............................27 Table 4.4 Pearson’s r value for each CRS Series ..............................................................28

vi

LIST OF FIGURES Figure 3.1 Study Region ....................................................................................................17 Figure 3.2 Dasymetric Example ........................................................................................21 Figure 3.3 Dasymetric Method Applied ............................................................................22 Figure 3.4 Block Group Selection Flow Diagram .............................................................22

vii

CHAPTER 11 INTRODUCTION Flood-prone communities in the United States are experiencing increasing threats from flooding events associated with climate change. These threats are associated with both intensified storm events and changing long-term sea levels influencing tidal flooding (Sweet and Park 2014; Moftakhari et al. 2015). Homeowners with federally-backed mortgages within the 100-year floodplain of these communities must participate in the National Flood Insurance Program (NFIP), which provides flood insurance to aid in personal recovery after a flooding event. Incentives and policies within the NFIP help communities’ recovery after a flooding event and place an emphasis on reducing and preventing flood losses. In recent years, the NFIP has come under some scrutiny due to mounting debt from large flooding events (e.g., Hurricanes Katrina and Ike, Super Storm Sandy). This attention encouraged efforts to make insurance premiums reflect actual risk exposure and reduce communities’ exposure to flood risks (Bellomo et al. 1999; King 2013; Knowles and Kunreuther 2014; Kousky and Kunreuther 2014; Kousky and Shabman, Pricing Flood Insurance: How and Why the NFIP differs from a Private Insurance Company, unpublished report, 2014; Michel-Kerjan 2010; Michel-Kerjan et al. 2012; NRC 2015). The NFIP’s untenable position has led researchers to examine other facets of the program to determine the best ways to improve the program (King 2013; Knowles and Kunreuther 2014; Kousky 2010; NRC 2015). 1

Landis, Z. P. To be submitted to Natural Hazards Review. 1

1 .1 BACKGROUND Within the NFIP exists the Community Rating System (CRS), which allows communities to earn reductions in NFIP insurance premiums through voluntary participation in hazard reduction activities. The CRS contains four series of activities: public information activities; mapping and regulations; flood damage reduction activities; and, warning and response. Each series contains several activities. Table 1.1 summarizes the CRS series, activities, maximum points per activity and presents descriptive statistics on the level of community participation by activity within the study region. For comparison, Table 1.2 provides the national statistics for CRS. Generalizing the comparison between the national CRS and the study area CRS, within the study area communities tend to have higher participation in all the activities but lower average scores for each activity. Table 1.1 2013 to Present: Study Region CRS Points by Series and Activity.

Activity

310 Elevation Certificates 320 Map Information Service 330 Outreach Projects 340 Hazard Disclosure 350 Flood Protection Information 360 Flood Protection

Maximum Possible Points

GA, NC, SC Maximum Points Earned

GA, NC, SC Average Points Earned

300 Series: Public Information Activities 116 116 67

GA, NC, SC Percentage of Communities Credited 100%

90

140

134

96%

350

310

158

91%

80

80

10

98%

125

102

57

97%

110

70

21

37%

2

Assistance 370 Flood Insurance Promotion * 410 Floodplain Mapping 420 Open Space Preservation 430 Higher Regulatory Standards 440 Flood Data Maintenance 450 Stormwater Management 510 Floodplain Mgmt. Planning 520 Acquisition and Relocation 530 Flood Protection 540 Drainage System Maintenance

110

0

0

400 Series: Mapping and Regulations 802 115 25

0%

94%

2,020

454

93

80%

2,042

914

297

100%

222

213

108

96%

755

370

65

93%

500 Series: Flood Damage Reduction Activities 622 310 107

66%

2,250

300

13

14%

1,600

482

13

11%

570

330

167

93%

600 Series: Warning and Response 610 Flood 395 205 74 Warning and Response 620 Levees * 235 0 0 630 Dams * 160 64 50 Note: * denotes new activity as of 2013 (based on FEMA 2013)

61%

0% 100%

Table 1.2 2013 to present: National CRS Points by Series and Activity.

Activity

Maximum Possible Points

National Maximum Points Earned

National Average Points Earned

300 Series: Public Information Activities

3

National Percentage of Communities Credited

310 Elevation Certificates 320 Map Information Service 330 Outreach Projects 340 Hazard Disclosure 350 Flood Protection Information 360 Flood Protection Assistance 370 Flood Insurance Promotion * 410 Floodplain Mapping 420 Open Space Preservation 430 Higher Regulatory Standards 440 Flood Data Maintenance 450 Stormwater Management 510 Floodplain Mgmt. Planning 520 Acquisition and Relocation 530 Flood Protection 540 Drainage System Maintenance

116

116

45

100%

90

70

50

93%

350

175

72

89%

80

57

19

71%

125

98

39

92%

110

65

49

41%

110

0

0

0%

400 Series: Mapping and Regulations 802 585 64

50%

2,020

1,548

463

70%

2,042

784

213

99%

222

171

87

89%

755

540

107

84%

500 Series: Flood Damage Reduction Activities 622 273 167

46%

2,250

1,701

165

24%

1,600

632

45

12%

570

449

212

77%

600 Series: Warning and Response 4

610 Flood 395 353 129 Warning and Response 620 Levees * 235 0 0 630 Dams * 160 0 0 Note: * denotes new activity as of 2013 (based on FEMA 2013)

37%

0% 0%

The CRS offers communities the opportunity to implement adaptive measures and build adaptive capacity in the community (A. Atreya, An Assessment of the National Flood Insurance Program’s (NFIP) Community Rating System (CRS), presented at 2016 Fall Conference: The Role of Research in Making Government More Effective, 2016; Atreya and Kunreuther, Measuring Community Resilience: The Role of the Community Rating System (CRS), unpublished report, 2016; Highfield and Brody 2017; Smit and Wandel 2006). In terms of flooding, this means the CRS is able to help reduce flood losses in the community and increase the community’s ability to respond to a flooding event (Kousky 2010). This loss-reduction and response-improvement potential makes the CRS a valuable program within the NFIP. As Table 1.1 shows, there are differences in the degree of community participation in each activity and series. For some activities, most communities are able to access a large percentage of the total points available. In other activities, communities earn on average only a small portion of the points available (e.g., communities average for floodplain mapping is 25 out of 802 possible points or 3%). While this flexibility in the point strategy allows communities to participate in activities that appeal to their citizens, it can also limit loss-reduction of the CRS by allowing communities to select activities with limited loss-reduction benefits. Brody et al. (2009b) found this very effect in Florida communities who all tended to earn more points in public information activities, which have limited loss reduction potential. Many of the activities available

5

under warning and response were not widely utilized because several of the activities are tied to state led programs (Brody 2009a). In part to improve the incentive system, the CRS undergoes strategic revisions approximately every five years with the most recent change occurring in 2013. These revisions may add activities and change the points associated with activities (e.g., the points for activity 420 Open Space Preservation changed from 900 in the 2007 CRS Manual to 2020 in the 2013 CRS Manual). The 2013 point system (Table 1.1) delegated more points towards activities that demonstrated greater benefits in reducing flood losses and damages (FEMA 2007, 2013). The 2013 point system places communities within Special Flood Hazard Areas (SFHA) in classes 10 to 1, providing reductions of 5% per class from (0% to 45% correspondingly). Communities outside of SFHAs are eligible for a 5% reduction only. Table 1.3 outlines the insurance premium reductions and the points associated with each CRS class. Table 1.3 CRS Classes CRS Class

CRS Credit Points

1 2 3 4 5 6 7 8 9 10 Source: FEMA 2013

4,500+ 4,000 – 4,499 3,500 – 3,999 3,000 – 3,499 2,500 – 2,999 2,000 – 2,499 1,500 – 1,999 1,000 – 1,499 500 – 999 0 – 499

Premium Reduction In SFHA Outside SFHA 45% 5% 40% 5% 35% 5% 30% 5% 25% 5% 20% 5% 15% 5% 10% 5% 5% 5% 0 0

In 2013, there were several major revisions to the CRS. Within the 400 series the total points stayed nearly the same, but the allocation of points within the series changed 6

significantly. Open space preservation (activity 420) jumped from a maximum of 900 points pre-2013 to 2020 points in the 2013 version (FEMA 2007, 2013). This change is important as that specific activity is credited with significantly reducing flood losses in communities (Brody 2012). The 300, 500, and 600 series also experienced changes, with the largest change affecting the 600 series. The change to the 600 series dropped its total possible points from 1330 pre-2013 to 790 points in the 2013 version (FEMA 2007, 2013). This change may be less significant within the program due to the limited number of communities accessing points within the 600 series (Table 1.2). These changes in the point system could influence which socioeconomic variables are the strongest correlates to higher levels of CRS participation. Although the CRS program is receiving considerable interest, the level of participation among communities is uneven. Research to investigate which community characteristics correspond most closely with higher levels of participation has begun. However, research to date has not fully explored various methods and approaches to verify findings. In addition, changes to census areas (i.e., incorporated towns annexing lands), advances in GIS (Geographic Information System) technology, and changes to the CRS point system could affect previous findings. This research uses updated community participation information, incorporates the use of an index, rather than a single value, and employs a dasymetric approach to identify the population and their characteristics within the floodplain validate previously identified indicators of CRS participation. 1.2 RESEARCH QUESTIONS This research aims to answer three questions regarding community CRS participation. In light of revisions to the CRS in 2013, 1) do previously identified

7

indicators of community CRS participation remain useful, for overall points and for points within each CRS series? 2) Are there significant differences between SoVI® and single community factors (educational attainment, housing value, population density) in the correlation to the CRS points? 3) How does the application of a dasymetric approach to identify population characteristics within the 100-year floodplain compare to the correlations of SoVI® and individual factors (educational attainment, housing value, population density) with community CRS participation calculated at the community level?

8

CHAPTER 2 LITERATURE REVIEW Researchers have focused on different programmatic aspects, areas, and community types in their CRS examinations. There are two common questions and methods in the current body of research (Brody et al. 2009b; Brody and Highfield 2013; Highfield and Brody 2012; Landry and Li 2012; Posey 2009; Sadiq and Noonan 2015a, b; Zahran et al. 2008a, b, 2010). The first common question investigates socio-economic factors associated with greater levels of overall CRS participation and the second considers factors associated with greater participation in each individual CRS activity (Table 1.1). Investigations of these questions generally share two methodological practices. They examine correlations of single census variables with CRS class, total points scores, and series points scores. The analyses also rely on data representing the entire community, as designated in CRS participation. The majority of these studies offer insight into community participation in the CRS program prior to the 2013 revisions to the point system, leaving open the question of how changes to the point system may have influenced patterns of community participation. Researchers have drawn from census variables to identify individual socioeconomic factors that indicate higher participation in the CRS overall and within each of the CRS series. Researchers relied on correlation analyses to conduct these studies, using specific techniques appropriate to the nature of their data (Brody et al. 2009b; Brody and

9

Highfield 2013; Highfield and Brody 2012; Landry and Li 2012; Posey 2009; Sadiq and Noonan 2015a, b; Zahran et al. 2010). In these analyses, researchers examined correlations with CRS classes or CRS points (Table 1.3). The total CRS points approach is stronger than examining CRS participation using the CRS classes because it fully captures the degree of participation. There are potentially points and related actions not captured by using CRS class (i.e., a community may earn up to 499 points over the threshold for given class, but still remain within that class), but it does increase the overall total CRS score. Using Pearson’s r, instead of Logit, Probit or Cragg models, provides an easily replicated method for future researchers to examine. The four major studies engaging community CRS participation identified socioeconomic factors that correlate to different measures of community CRS participation. The methods, regions and community types in these studies differ slightly. There are both some consistencies and differences among the findings. Following brief descriptions of each, Table 2.1 summarizes key aspects of these studies. Brody et al. (2009b) aimed to determine how Florida counties participated in each of the CRS series. They determined educational attainment had a positive correlation with community participation in each of the CRS series while population density had no correlation. This study did not assess housing value (Brody et al. 2009b). This study provides the only insights into specific CRS series, illuminating the importance for examining specific CRS series. Landry and Li (2012) examined CRS participation in North Carolina for counties only, examining if competing factors (e.g. crimes rates, educational system) limited some counties’ CRS participation. Their study found population density and housing value to

10

have positive correlations with CRS participation. They also identified a negative correlation between educational attainment and CRS participation. Posey (2009) aimed to determine adaptive capacity of municipalities throughout the US and used the CRS class as a proxy to measure adaptive capacity. He identified both educational attainment and housing value to have statically significant positive correlations to CRS participation. There was not a population density factor included in this analysis. Sadiq and Noonan (2015b) examined all CRS communities in the US attempting to provide definitive indicators nationally for both CRS participation and CRS score, foregoing examining each CRS series or activity. They identified both educational attainment and housing value to have positive correlations to CRS participation, educational attainment to have a positive correlation to CRS score and no correlation to either participation or score for population density. Table 2.1 Summary of Related Research Study

Time

Brody et al. (2009b) Landry and Li (2012)

1999 – 2005 1991 – 2002 1978 2007

Posey (2009) Sadiq and Noonan (2015)

2012

Region

Enum Unit

N

Stat Method

Independ. Variables

Dependent Variables

Florida

Counties

48-51

Multivariate stat models

Census Variables

CRS Series and Activities

North Carolina

Counties

100

Logit and Probit Models

Census Variables

CRS Participation

National

Municipalities

10,916

Probit and OLS Models

Census Variables

CRS Participation

National

All CRS Commun ities

1,182

Two-stage Cragg model

Census Variables

CRS Participation and Score

Note: Enum Unit = Enumeration Unit; N = sample size; Stat Method = Statistical Method Used

11

Using these methods, research identified educational attainment, home value, and population density as indicators of higher CRS participation; however, findings were not consistent across studies (Brody et al. 2009b; Posey 2009; Landry and Li 2012; Sadiq and Noonan 2015b). Educational attainment is defined as percentage of the population (25 and older) with a bachelor degree or higher; home value is represented by median home value; population density refers to population per square mile. Landry found no correlation between CRS and educational attainment while Brody et al., Posey, and Sadiq and Noonan identified educational attainment as a correlate. Landry and Li’s (2012) investigation of counties in North Carolina was the only study to identify population density as a strong indicator of higher participation. Table 2.2 summarizes the findings of these studies relating selected census variables to CRS participation. These specific variables were chosen for several reasons: educational attainment has consistently some of the highest correlations to CRS scores, educational attainment and housing value were consistently found to have significant correlation. Only one study found population density as a correlate. Table 2.2 Studies Showing Census Factor Correlation to Aspects of the CRS.

Brody et al. (2009b) Landry and Li (2012) Posey (2009) Sadiq and Noonan (2015b)

Educational Attainment +

Housing Value

Population Density

N/A





+

+

+ +

+ +

N/A ≠

12

A significant difference in these studies lies in their spatial extent and region examined. The Posey (2009) and Sadiq and Noonan (2015b) studies are national studies examining CRS communities while Brody et al.’s (2009b) study focused on communities in Florida. Landry and Li (2012) examined communities in North Carolina. Sadiq and Noonan (2015b) examine all CRS communities (counties and incorporated towns, riverine and coastal) within their study region, while Brody et al. (2009b) and Landry and Li (2012) examined only counties in their study areas. Posey (2009) focused on municipalities (incorporated places smaller than county). These differences in community type (e.g., county or smaller municipality) and region (e.g., Colorado or Georgia) offer one potential explanation for the differences among findings. Another issue relevant to understanding community CRS participation is that communities’ participation differs across series and for each activity of the series. Sadiq and Noonan (2015b) call for an examination of these correlations to each CRS series to determine if overall CRS predictors show similar patterns for each of the CRS series. Brody et al. (2009b) and Brody and Highfield (2013) highlight the importance of investigating differences in participation within each series. Communities in Florida are more likely to participate in informational campaigns (i.e. 300 series activities) than in other CRS series, which require more effort from the community (Brody et al. 2009b). Brody and Highfield (2013) also demonstrates how open space preservation (an activity in the 400 series) is linked to flood loss reduction. These findings are important distinctions within the CRS, demonstrating the activities that result in the greatest reduction of losses and the activities communities elect to pursue are not always aligned. One caveat for the CRS series and activity evidentiary basis for this research is that is

13

much of it is based in the efforts of a few researchers collaborating with Brody in many studies. Each series in the CRS has a different focus and each requires varying amounts of political and fiscal capital and other resources to enact. The 300 series tends to contain informational campaign activities; the 400 series contains data archiving, regulatory changes, and mapping; the 500 series focuses on flood reduction activities; the 600 series addresses warning systems and dam and levee maintenance. The two right columns of Table 1.1 demonstrates the varying participation in each series and shows the average achievement communities attain in that activity within the study region. One of the most consistent approaches across these studies is the use of single socio-economic variables as a correlation to CRS scores (Brody et al. 2009a, b; Brody and Highfield 2013; Highfield and Brody 2012; Landry and Li 2012, Landry and JahnParvar 2011; Sadiq and Noonan 2015a, b; Zahran et al. 2008a, 2010). However, the use of composite measures, such as indices and scales, practiced in other fields, could consolidate several socio-economic factors to provide a better correlation than a single factor (Babbie 2013). Capturing the combined influence of multiple variables through the use of an index may provide greater explanatory power in understanding the relationships between community characteristics and CRS participation. In the area of hazard management, understanding vulnerability of a community is a particularly important component in understanding overall risk. Cutter et al. (2003) provides just such a hazard relevant index through the Social Vulnerability Index (SoVI®). SoVI® reduced 42 socio-economic variables through principal components analysis into 11 factors, which explained 76% of the variance in the

14

single factors. Educational attainment and housing value may each capture some aspect of vulnerability, so testing SoVI® as a possible indicator for CRS score could provide greater insights into participation among CRS communities as it represents a more holistic measure of vulnerability. The second consistent research approach is utilizing data summarizing the entire community’s socio-economic background (Brody et al. 2009a, b; Brody and Highfield 2013; Highfield and Brody 2012; Landry and Li 2012, Landry and Jahn-Parvar 2011; Sadiq and Noonan 2015a, b; Zahran et al. 2008a, 2010). These researchers examined the community as a whole, rather than examining only individuals who live in the floodplain and are thus more affected by changes in CRS score (e.g., increased flood protection and decreased insurance rates). Communities participating in the CRS vary in both population and land area. Differences in residential patterns in these communities include the level of floodplain occupancy. For example, Hilton Head, SC (pop. 37,000 and 37 sq. mi.) is almost entirely in the 100-year floodplain, while Jacksonville, NC (pop. 70,000 and 46 sq. mi.) only has small sections of the community in the 100-year floodplain. This variation in the occupancy in the floodplain could potentially produce significant changes in CRS involvement in the community. In situations, where the precise location of residences with respect to some boundary is critical to the analysis, researchers have frequently applied a dasymetric approach (Langford and Higgs 2006). Briefly, a dasymetric approach takes into account the changing densities of population within the boundaries of the map. Given the potential for population characteristics to influence patterns of investment within a

15

community, it is possible that a dasymmetric approach may reveal different relationships between indicators and CRS scores.

16

CHAPTER 3 METHODS This study includes all CRS communities in the coastal counties (as each state defines them under the Coastal Zone Management Act) of North Carolina, South Carolina, and Georgia (n=88) (Appendix A). Figure 3.1 shows the study region for this research. Each state has slightly different definition of their coastal counties. North Carolina’s definition is the most unlike South Carolina or Georgia. North Carolina coastal counties are only those counties that border coast or coastal sound, while both South Carolina and Georgia include counties effected by tidal waters. This research did not identify any communities in this region which were eligible for the CRS but chose not to participate. There were no communities identified that had opted out of the NFIP.

Figure 3.1 Study Region (Outlined in black). The study region is 20 counties from North Carolina, 8 counties from South Carolina and 11 counties from Georgia. The presentation of methodology begins with a description and justification of data sources and associated caveats for analysis. A discussion of statistical techniques employed follows. Details of the dasymetric approach used are addressed in the final 17

section. This research uses floodplain data from the 2016 National Flood Hazard Layer, CRS score data from 2014 FEMA State profiles, census data from the 2013 ACS 5-year estimates, and 2000 SoVI® data from NOAA’s Digital Coast (FEMA 2016; Census 2013; NOAA 2000). The 2016 National Flood Hazard Layer contains all floodplains within the United States. This research used the 100-year floodplain as designated by the National Flood Hazards Layer. The CRS score data was provided from the 2014 FEMA State profiles. This data (Appendix B) included every CRS community in North Carolina, South Carolina and Georgia with the CRS class each community attained as well as the community’s overall score, score for each series and score for each activity. The assessment of the characteristics of floodplain occupancy relied on community block group data from the 2013 ACS 5-year estimate (2009-2013). These data provided the lowest margin of error at the block group level. For assessments of the entire community, 2013 ACS 5-year estimates (2009-2013) for census place (county or incorporated town) were used as the margins of error associated with these data were much lower than the block group data for 1-year and 3-year estimates. The margin of error for the census place was under 10% of the mean where the margin of error for the block groups ranged from under 10% of the mean to 60% of the mean. Educational attainment is defined as percentage of the population (25 and older) with a bachelor degree or higher; home value is represented by median home value; population density refers to population per square mile. NOAA’s Digital Coast provided year 2000 SoVI® data at the block group level. Using these data allows for the analysis of the SoVI® at the scale of the floodplain in each community.

18

Two caveats apply to the data used. The margin of error associated with the block group data could be an item of concern. However, as the results of this analysis show, there is no statistical difference between the floodplain data and the community data meaning examination of just the population within the floodplain provides no additional insight to CRS score participation. If the results had demonstrated a statistical difference indicating the floodplain provided more accurate representation of the community more attention to the influence of the margin of error would be warranted. The SoVI® data is from year 2000, which does not provide the most accurate picture of communities in 2013. This data is the most recent available at the smallest census unit because this research was unable to more recent SoVI® data. While not a close match in the time period, it is sufficient to provide preliminary insights into the use of an index in examining CRS communities. In order to compare single census factors-CRS point score correlations with one another, the correlations are examined using Pearson’s r and transformed using Fisher’s z transformation. Pearson’s r is a measure of linear correlation with values ranging from total negative to total positive correlation, -1 to 1 respectively. Fisher’s z transformation changes the values from r to normally distributed z-scores allowing for a more accurate comparison of correlation values (Mudholkar 2004). All reported Pearson’s r values in the results are compared using Fisher’s z transformation. The following example demonstrates the value in Fisher’s z transformation, if you compare the r values of .95 to .90 and .55 to .50, they seem to have the same difference .05 however, Pearson’s r only has a range of -1 to 1, so as you approach either end of the value range small changes are statistically significant. Taking the r values provided above, there is a significant

19

statistical difference between r values of .95 to .90 where there is no statistical difference between the r values of .55 to .50. The linear approach is acceptable because the correlation is to continuous CRS points score, rather than ordinal CRS classes. Using the three socio-economic factors identified from the literature (educational attainment, housing value, and population density), this research uses the same correlation analyses, described in the previous paragraph, to examine which variables have the highest correlations with CRS series. This analysis follows Brody (2009b, 2013) to determine if the CRS changes in 2013 brought about changes in the type of community that attempts to access each CRS series and Sadiq and Noonan’s (2015b) recommendations for determining which indicators are strongest for each CRS series. Using an index allows for the consolidation of several factors into one component, these components tend to group around factors that have similar high and low correlations. The assessment of the index value rather than a single factor is conducted using the SoVI® (Cutter et al. 2003). The reduction of single variables into an index allows for a better understanding of specific communities by aligning factors under similar themes. For example, the index could place housing value, educational attainment, and median income into a single factor that the researcher could then call their level of affluence factor. The SoVI® is widely used in hazards research applications (Lam et al. 2016; Posey 2009). Because of the broad application and focus on hazards, SoVI® is a good choice for investigation of the potential for indices. This research employs a dasymetric approach to increase the spatial accuracy of population data relative to the floodplain. Previous CRS research used an equal distribution method, where the data is evenly distributed throughout the spatial unit

20

(Brody et al. 2009b, b; Brody and Highfield 2013; Highfield and Brody 2012; Landry and Li 2012, Landry and Jahn-Parvar 2011; Sadiq and Noonan 2015a, b; Zahran et al. 2008a, 2010). In contrast, dasymetric mapping places the data points as close to their real-world location as possible. Figure 3.2 provides a general visualization of the difference between using a single point (left), what previous research has done (center), and what this research does (right).

Figure 3.2 Dasymetric Example. The three alternative representations of population used in accessibility modeling: using a single representative point (left), evenly distributed throughout a census zone (center), and a dasymetrically distributed model (right) (Langford and Higgs 2006). Figure 3.3 displays the specific manner in which this research applied the dasymetric approach. The large box represents the whole community and the grey shaded area represents the floodplain within the community. The checkered boxes within the whole community represent smaller census units (block groups). As the graphic depicts, there are still some inaccuracies with this approach. The census units do not perfectly line up with the floodplain, meaning census units were used to describe the demographics of the floodplain that were partially outside the floodplain and some census units were not used that were partially inside the floodplain.

21

Figure 3.3 Dasymetric Method Applied. Example of the dasymetric approach used in this research. Outer box is the entire community, grey shaded area is the floodplain in the community, and checkered boxes are census block groups with their areal centroid in the floodplain. Figure 3.4 demonstrates the specific manner block groups were selected for the dasymetric approach. This research evaluated three different methods for selecting block groups in the community’s floodplain. The first was selecting block groups that intersected the community’s floodplain. This resulted in overestimation as several communities had larger populations and land area included in the floodplain. The second method selected block groups completely contained within the community’s floodplain. This method resulted in underestimations and some of the smaller communities had no demographic information for floodplain areas. The last method considered and the one used for this research was selecting block groups with the centriod within the floodplain. Figure 3.4 represents the flow of this block group selection process.

Figure 3.4 Block Group Selection Flow Diagram. This figure demonstrates the workflow of block group selection for the dasymetric approach applied. 22

Utilizing more precise location data can change results (Landford and Higgs 2006; Openshaw 1983). Using this dasymetric approach in examining a community’s CRS score allows the research to focus more closely on the population in the floodplain and largely remove those population in the community unaffected by changes in the CRS. This research used paired t-tests to determine if there was a significant difference between the data at the whole community and the floodplain levels (Hsu and Lachenbruch 2008). The specific application of the dasymetric method was accomplished through ArcGIS, selecting census data by location (inside the community and inside the floodplain). There are still limits to this dasymetric approach, which encounters the modifiable areal unit problem (MAUP), albeit on a smaller scale than previous studies. With improved data representation this problem could be overcome completely (Holt et al. 2004, Manntay et al. 2007).

23

CHAPTER 4 RESULTS The results from this study provide insights into CRS participation by confirming previous findings, applying SoVI® to examine CRS communities, examining CRS series participation, and evaluating influence of community socio-economic composition through dasymetric analysis. The results help resolve some differences in past research on CRS participation indicators and offer preliminary insights into examining data dasymetrically. 4.1 TOTAL CRS SCORE AND DEMOGRAPHIC ASSOCIATIONS Consistent with the findings from previous research, housing value and educational attainment remain strong indictors of overall CRS score. Table 4.1 lists the results of the Pearson’s r analysis for the factors and locations studied. North Carolina communities are included in Table 4.1 to facilitate comparison with Landry and Li (2012). The positive correlation between housing value and CRS score and educational attainment and CRS score demonstrates that as housing values and education attainment in a community increase the communities CRS score will likely increase. Table 4.2 lists the specific significance levels for r values of with sample sizes of 48 and 88. The values in Table 4.2 demonstrate that housing value, educational attainment and SoVI® score are significant correlations above 99% confidence. The table shows this same significance

24

demonstrated in the whole study region applies for North Carolina communities and demonstrates no correlation between CRS scores and population density. Table 4.1 Correlation Values (Pearson’s r) Between Indicator and Total CRS Point Scores. NC Coastal Communities (N=48) 0.59*** Median Housing Value 0.54*** Educational Attainment 0.14 Population Density ® SoVI Score -0.49*** Note: *** statistically significant at p<0.01

NC, SC, GA Communities (N=88) 0.44*** 0.47*** 0.01 -0.39***

NC, SC, GA Floodplains (N=88) 0.43*** 0.46*** -0.01 -0.43***

Table 4.2 Statistically Significant r Values. N

df

0.1 (90%) 0.997 0.950 0.878 0.811

0.05 (95%) 0.999 0.975 0.924 0.868

3 1 4 2 5 3 6 4 … … 48 46 0.285 0.323 … … 88 86 0.210 0.239 Note: α level for two-tailed test (Rogerson 2001)

0.02 (98%) 1.000 0.990 0.959 0.917

0.01 (99%) 1.000 0.995 0.974 0.942

0.368

0.399

0.273

0.297

Interestingly, the SoVI® score proved to be an indicator of total CRS score as well. The SoVI® score includes measures of educational attainment and household income as well as other variables, such as renter and elderly populations, that capture more dimensions of community vulnerability. A SoVI® score of -5 would mean the area has very low social vulnerability. This correlation means the lower the social vulnerability in an area the higher the CRS score. It also means that locations with higher

25

social vulnerability appear to not participate as much in a program that would help lower their exposure to flooding events. Unlike Landry and Li’s (2012) findings, this research did not find population density to be a indicator to overall CRS points scores (Table 4.1). The difference between the findings likely stems from the difference in types of communities included in the studies. Landry and Li only examined counties in North Carolina, leaving out smaller CRS communities. This study examined all CRS communities in the coastal zones of three states, including the municipalities below county level. However, when North Carolina communities were isolated (n=48), this study was still unable to reproduce Landry and Li’s findings. While the correlation value for population density did increase, using Fisher’s z transformation to test the statistical difference between population density and the other indicators showed that these are significantly different above a 95% confidence threshold. That analysis indicates that housing value, educational attainment, and SoVI® score are all statistically better indicators of CRS participation than population density in both the whole study area and in North Carolina. Furthermore, Table 4.2 shows the r value associated with 90% certainty is 0.285, substantially above the r value for North Carolina’s population density (r = .14), indicating that population density does not indicate CRS participation. The dasymetric mapping approach facilitates a more precise location of population in a study area. This approach allows determination of statistical difference between the floodplain and the entire community. When conducting a paired t-test between the floodplain and whole community, housing value and population density are shown to have a significant statistical difference (Table 4.3). Since population density is

26

shown not to correlate with CRS score, the statistical difference only demonstrates how values can differ when located specifically. Similarly, the difference between housing value in the floodplain and the whole community shows a statistical difference; however, when comparing the r values (using Fisher’s z transformation), there is no statistical difference between the floodplain and the whole community. This analysis means the applied dasymetric approach adds limited value to further understanding CRS communities in this study area. This dasymetric approach did not provide a distinction between the floodplain and the whole community. The aim of this dasymetric approach was to demonstrate that those living in the floodplain would have more of an impact on CRS score indicators than the whole community. Table 4.3 Paired T-Test between Floodplain and Whole Community Mean Floodplain Value 297,001

Median Housing Value 33.8% Educational Attainment 434 Population Density ® SoVI Score -1.12 Note: * Statistical difference p value = 0.01

Mean Community Value 279,413

Two-tail T-test p Value

33.3%

0.34479

628 -1.08

0.0000002*** 0.74652

0.00008***

4.2 SERIES CRS SCORE AND DEMOGRAPHIC ASSOCIATIONS Brody et al. (2009b) and Brody and Highfield (2013) examined community participation to specific series within the CRS. They demonstrated how communities choose to participate in CRS series and which series reduce flood losses. These results examine indicators of CRS participation by series, highlighting the difference between the floodplain and the whole community (Table 4.4). Fisher’s z transformation demonstrates the only statistical differences are those between population density and the

27

other variables, meaning while educational attainment in the community consistently has the highest r values across the CRS series, those r values are not statistically larger than housing value or SoVI® score. This means, statistically speaking, housing value and SoVI® score are as good as educational attainment at indicating CRS participation by series. Table 4.4 Pearson’s r value for each CRS Series (n=88). Highlighted values show the highest correlation to each series. Housing Education PopDen SoVI® H-FP E-FP P-FP S-FP 0.392 0.357 -0.058 -0.235 0.370 0.379 -0.054 -0.302 300 *** *** *** *** ** 0.279 0.347 -0.018 -0.384 0.297 0.347 -0.089 -0.401 400 ** *** *** *** *** ** 0.323 0.161 -0.144 0.303 0.321 0.133 -0.133 0.282 500 *** *** *** ** 0.336 0.322 0.042 -0.242 0.309 0.298 -0.046 -0.293 600 *** *** * *** ** Note: H = Housing Value; E = Educational Attainment; P = Population Density; S = SoVI®; FP = Floodplain. * p<0.05; ** p<0.02; ***p <0.01 These correlations to CRS series suggest that, despite Brody and Highfield’s 2012 findings demonstrating activities in the 400 series perform best at reducing flood losses, communities continue to pursue 300 series more heavily. Table 4.4 demonstrates this by the higher correlation values for 300 series, these results taken in tandem with the results for total CRS score (Table 4.1) demonstrate the more points received from a specific series the higher the r value for that series. Brody et al. (2009b) also found communities participating at higher levels in the 300 series. There are several possible implications for this pattern: one is that communities will pursue the less resource-intensive points first (300 series) and build consensus to address more resource-intensive activities in the 400 and 500 series. Table 1.1 shows 300 series has high participation and high awarding of possible points; the 400 series has high participation and lower awarding of possible

28

points; and, the 500 series has lower participation and lower awarding of points. A second possibility is that communities do not necessarily prioritize reducing flood losses, and focus on the more accessible activities to reduce their insurance premiums.

29

CHAPTER 5 DISCUSSION The results from this research provide three key insights into understanding community participation in the CRS. The major takeaway from this research is the correlation of the SoVI® scores to CRS participation. This research demonstrates that overall vulnerability is associated with CRS participation; the lower the vulnerability the higher the participation and vice versa. Possibly the most troubling finding from this investigation of SoVI® is the indication that the more social vulnerability a community experiences, the less likely it is to participate at higher classes of CRS. Meaning, the most vulnerable populations are not participating equally in measures that could lower their vulnerability to flooding hazards. The SoVI® provides a conceptually stronger approach for representing the relationship between vulnerability and CRS scores. Despite SoVI® having slightly lower r values than both housing value and educational attainment, the Fisher’s z transformation indicates that the correlations are not significantly different (two tailed pvalue between SoVI® and housing value is .69 and two tailed p-value between SoVI® and educational attainment is .52). This strength stems from the multiple factors SoVI® considers versus only the single factors. This multi-factored approach provides a more complex understanding of the vulnerability of participating communities by accounting for more of the demographic characteristics associated with community vulnerability. In more practical terms, this means that while the single factor only says one thing about the

30

community, SoVI® suggests that communities who participate more in CRS tend to have a combination of variables associated with lower vulnerability, such as higher housing values, higher mean incomes, larger Asian populations, higher educational attainment, and lower renter populations. This previous illustration is only for practical understanding of SoVI®, as this research only used total SoVI® scores and not the individual drivers to SoVI® it cannot make claims regarding individual drivers. These findings also show that despite changes to the CRS in 2013, housing value and educational attainment continue to correlate with higher levels of CRS participation. The persistence of the correlations between CRS scores, educational attainment and housing value suggests that regardless of changes to the CRS program, communities who are participating in the program will continue to be involved. This finding may indicate the need to consider a programmatic shift that better incentivizes all types of communities to participate in the CRS. Landy and Li’s (2012) finding that population density is an indicator of CRS participation is not confirmed by this study. Based on these correlations, coupled with the findings about SoVI®, the CRS appears to be a program with less engagement from less affluent and more socially vulnerable communities. In other words, communities which, arguably, could benefit the most from the CRS appear to be participating the least. The final point is that a dasymetric approach provides limited value for understanding CRS communities in these NC, SC, and GA coastal zones. While the dasymetric approach demonstrated a better correlation in determining 400 series participation, the difference between the floodplain and the whole community was not statistically significant. A possible reason for the limited benefits of examining CRS

31

communities dasymetrically is the selected study region. Most of the communities in the study region had sections that were not in the floodplain; over half of the communities had more than 50% of the community in the floodplain. With such a large portion of the community in the floodplain it is likely that the community as a whole is interested in changes to their community CRS score. A more definitive approach for examining CRS communities dasymetrically may be to study CRS communities in riverine floodplains where larger sections of the community will be outside the floodplain or by using parcellevel data to further refine the population’s location.

32

CHAPTER 6 CONCLUSION The NFIP is designed to reduce community members’ exposure to flood losses, allowing individuals to receive flood insurance. Within the NFIP the CRS allows communities to take collective loss-reducing actions, which in turn lower the communities’ insurance premiums and protect the community from flooding events. Understanding which communities participate in the CRS can assist program administrators in refining the CRS to reduce adaptation barriers some communities may experience in participating in the CRS (Klein et al. 2014). Increasing community participation in the CRS would assist in reducing flood losses, which could also reduce post-flood recovery time. The alternative methods for examining CRS communities examined in this research allows for researchers to understand CRS communities through a new lens. This research demonstrated the value SoVI® in indicating community CRS points score, providing a more structurally sound indicator over the previous single-factors. The SoVI® proves to be an alternate indicator for researchers studying CRS communities, providing correlation coefficients consistent with the best single-factor indicators. This finding should encourage researchers to further examine this and other indices for their ability to provide additional insights into communities’ socio-economic standing and ability to respond to hazardous events. While previous research examined the whole CRS community, this research

33

examined CRS communities through a different approach considering the characteristics of floodplain residents. Past studies determined that flood risk exposure increased communities’ involvement in the CRS. This research followed those findings to determine if who specifically was exposed to that risk in the community affected how the community participated. This research showed there was no statistical difference between the whole community and the floodplain, suggesting that perhaps the community perception of flood risk exposure applies to the whole community and not just those in the floodplain. However, the limitations of the dasymetric approach applied here and the lack of information on flood risk perceptions require further research to explore that possibility. 6.1 FUTURE DIRECTIONS The CRS is a government program which truly addresses the purpose for which is was created: reducing flood losses in communities. The CRS needs to continue to reform to further address flood losses in communities and facilitate for a broader base of communities participating in the CRS. Two items come to mind for future revision to the CRS, the first is tiered requirements based on the number of years a community has been in the CRS. Over time, increasing the number of points required to maintain insurance discounts could push communities towards the more flood reducing activities in the CRS and further increase the effectiveness of the program. The second item to include is a socio-economic multiplier for communities. This socio-economic multiplier could be factored by considering the median income for the community, or the tax revenue for the community and if those are below a certain threshold the community would receive the multiplier. This multiplier would take the community’s overall score and multiply it by

34

the multiplier value (whatever that value is determined to be). While this would not initially reduce flood losses, it could encourage communities which do not have many resources to participate in the program. However, even with this multiplier, it would be important for communities to meet the tiered requirements described above after so many years in the CRS. This approach could allow less advantaged communities to access the CRS benefits without fundamentally changing the CRS to a social program. This access to the program could build consensus in the community allowing the community to build participation gradually.

35

REFERENCES Babbie, E (2013). Indexes, Scales, and Typologies. In The Practice of Social Research (pp. 158-174). Belmont, CA: Cengage Learning. Bellomo, D., Pajak, M. J., and Sparks, J. (1999). “Coastal flood hazards and the national flood insurance program.” Journal of Coastal Research, 21-26. Brody, S. D., and Highfield, W. E. (2013). “Open space protection and flood mitigation: A national study.” Land Use Policy, 32, 89–95. Brody, S. D., Kang, J. E., and Bernhardt, S. (2009a). “Identifying factors influencing flood mitigation at the local level in Texas and Florida: the role of organizational capacity.” Natural Hazards, 52(1), 167–184. Brody, S. D., Zahran, S., Highfield, W. E., Bernhardt, S. P., and Vedlitz, A. (2009b). “Policy Learning for Flood Mitigation: A Longitudinal Assessment of the Community Rating System in Florida.” Risk Analysis, 29(6), 912–929. Census (2013). “American Community Survey 5 year estimates.” (Dec. 28, 2016) Cutter, S. L., Boruff, B. J., and Shirley, W. L. (2003). “Social vulnerability to environmental hazards.” Social science quarterly, 84(2), 242-261. FEMA (2007). National Flood Insurance Program Community Rating System Coordinator’s Manual. Federal Emergency Management Agency, Washington, DC. FEMA (2013). National Flood Insurance Program Community Rating System Coordinator’s Manual. Federal Emergency Management Agency, Washington, DC. FEMA (2016). “National Flood Hazard Layer.” (Nov. 17, 2016) Highfield, W. E., and Brody, S. D. (2012). “Evaluating the effectiveness of local mitigation activities in reducing flood losses.” Natural Hazards Review, 10.1061/(ASCE)NH.1527-6996.0000114, 229–236.

36

Highfield, W. E., and Brody, S. D. (2017). “Determining the effects of the FEMA Community Rating System program on flood losses in the United States.” International Journal of Disaster Risk Reduction, 21, 396-404. Holt, J. B., Lo, C. P., and Hodler, T. W. (2004). “Dasymetric estimation of population density and areal interpolation of census data.” Cartography and Geographic Information Science, 31(2), 103-121. Hsu, H., and Lachenbruch, P. A. (2008). “Paired t test.” Wiley Encyclopedia of Clinical Trials, 1-3. King, R. O. (2013). “The national flood insurance program: Status and remaining issues for congress.” Congressional Research Service, 42850. Klein, R.J.T., G.F. Midgley, B.L. Preston, M. Alam, F.G.H. Berkhout, K. Dow, and M.R. Shaw (2014). Adaptation opportunities, constraints, and limits. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C.B., V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 899-943. Knowles, S. G., and Kunreuther, H. C. (2014). “Troubled waters: The national flood insurance program in historical perspective.” Journal of Policy History, 26(03), 327353. Kousky, C., and Kunreuther, H. (2014). “Addressing affordability in the national flood insurance program.” Journal of Extreme Events, 1(01), 1450001. Kousky, C. (2010). “Understanding the demand for flood insurance.” Natural Hazards Review, 10.1061/(ASCE)NH.1527-6996.0000025, 96-110. Lam, N. S., Reams, M., Li, K., Li, C., and Mata, L. P. (2015). “Measuring community resilience to coastal hazards along the Northern Gulf of Mexico.” Natural hazards review, 10.1061/(ASCE)NH.1527-6996.0000193, 04015013. Landry, C. E., and Jahan-Parvar, M. R. (2011). “Flood Insurance Coverage in the Coastal Zone.” The Journal of Risk and Insurance, 78(2), 361–388. Landry, C. E., and Li, J. (2012). “Participation in the Community Rating System of NFIP: Empirical Analysis of North Carolina Counties.” Natural Hazards Review, 10.1061/(ASCE)NH.1527-6996.0000073, 205–220.

37

Langford, M., and Higgs, G. (2006). “Measuring Potential Access to Primary Healthcare Services: The Influence of Alternative Spatial Representations of Population.” Professional Geographer, 58(3), 294–306. Maantay, J. A., Maroko, A. R., and Herrmann, C. (2007). “Mapping Population Distribution in the Urban Environment: The Cadastral-based Expert System (CEDS).” Cartography and Geographic Information Science, 34(2), 77–102. Michel-Kerjan, E. O. (2010). “Catastrophe economics: the national flood insurance program.” The Journal of Economic Perspectives, 24(4), 165-186. Michel‐Kerjan, E., Lemoyne de Forges, S., and Kunreuther, H. (2012). “Policy Tenure Under the US National Flood Insurance Program (NFIP).” Risk Analysis, 32(4), 644658. Moftakhari, H. R., AghaKouchak, A., Sanders, B. F., Feldman, D. L., Sweet, W., Matthew, R. A., and Luke, A. (2015). “Increased nuisance flooding along the coasts of the United States due to sea level rise: Past and future.” Geophysical Research Letters, 42(22), 9846-9852 Mudholkar, G. S. (2004). “Fisher's z‐Transformation.” Encyclopedia of Statistical Sciences. Wiley Online Library. National Research Council (2015). Affordability of National Flood Insurance Program Premiums: Report 1. Washington, DC: The National Academies Press. NOAA (2000). “Social Vulnerability Index Block Group Data.” (Dec. 29, 2016) Openshaw, S. (1983). The Modifiable Areal Unit Problem. Concepts and Techniques in Modern Geography No. 38. Norwich, England: Geobooks. Posey, J. (2009). “The determinants of vulnerability and adaptive capacity at the municipal level: Evidence from floodplain management programs in the United States.” Global Environmental Change, 19(4), 482-493. Rogerson, P (2001). Correlation. In Statistical Methods for Geography (pp. 211-213). Thousand Oaks, CA: Sage. Sadiq, A.-A., and Noonan, D. (2015a). “Local capacity and resilience to flooding: community responsiveness to the community ratings system program incentives.” Natural Hazards, 78(2), 1413–1428.

38

Sadiq, A.-A., and Noonan, D. S. (2015b). “Flood disaster management policy: an analysis of the United States Community Ratings System.” Journal of Natural Resources Policy Research, 7(1), 5–22. Smit, B., and Wandel, J. (2006). “Adaptation, adaptive capacity and vulnerability.” Global Environmental Change, 16(3), 282-292. Sweet, W. V., and Park, J. (2014). “From the extreme to the mean: Acceleration and tipping points of coastal inundation from sea level rise.” Earth's Future, 2(12), 579600. Zahran, S., Brody, S. D., Highfield, W. E., and Vedlitz, A. (2010). “Non-linear incentives, plan design, and flood mitigation: the case of the Federal Emergency Management Agency’s community rating system.” Journal of Environmental Planning and Management, 53(2), 219–239. Zahran, S., Brody, S. D., Peacock, W. G., Vedlitz, A., and Grover, H. (2008a). “Social vulnerability and the natural and built environment: a model of flood casualties in Texas.” Disasters, 32(4), 537–560. Zahran, S., Brody, S. D., Vedlitz, A., Grover, H., and Miller, C. (2008b). “Vulnerability and Capacity: Explaining Local Commitment to Climate-Change Policy.” Environment and Planning C: Government and Policy,

39

APPENDIX A – LIST OF CRS COMMUNITIES STUDIED

North Carolina Alliance Atlantic Beach Bayboro Beaufort Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville Kill Devils Hill Kitty Hawk Manteo Minnesott Beach Morehead City Nags Head

40

New Hanover County Newport North Topsail Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach Topsail Beach Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown Hilton Head Island Hollywood Horry County

41

Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia Brunswick Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

42

APPENDIX B – DATA USED

B.1 WHOLE COMMUNITY DATA

North Carolina Alliance Atlantic Beach Bayboro Beaufort Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville Kill Devils Hill Kitty Hawk

Median Housing Value ($) 128800 301600 78100 196500 86000 260300 258700 199200 462300 335800 152400 85700 223800 293900 573400 111100 390600 137600 481500 76400 153600 260900 321100

Educational Attainment (Percent Bachelors or higher 25 and over) 14.6 30.3 6.1 24 10.8 33.8 36.4 23.9 51.3 39 21 10.9 18.4 30.6 58.3 20.7 46.5 12.5 55.3 10.7 23.1 34.3 28.2

43

Population Density (per Sq Mi) 371.974243 642.7446929 681.7074889 871.4456733 1060.697428 762.4955332 2326.992668 130.9404778 135.4766843 579.6553681 146.4631885 487.6778419 89.89538272 88.51434466 152.7606505 931.8533665 733.2082145 1230.674105 212.1848841 9.488310847 1507.298829 1189.587161 403.3210281

SoVI® Score 2.36 -5.51 2.36 -3.95 2.9 -1.94 -5.03 -0.9 -3.17 -1.61 0.19 2.58 -2.15 -2.84 -5.1 2.3 -4.41 -3.18 -3.02 1.04 -2.7 -3.74 -4.66

Manteo Minnesott Beach Morehead City Nags Head New Hanover County Newport North Topsail Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach Topsail Beach Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown

400000 194400 191300 326200 215200 158700

32.3 35.8 27.5 38.6 36.6 9.2

757.5013891 126.7253444 1227.840426 418.8732251 1056.178366 563.9604527

-1.63 0.32 -0.02 -3.12 -0.87 -0.19

286800 243200 538700 322800 157500 393800 91400 193400 81800 455200 245800 57300 273100 391500 75300 89000 236600 159900 844700

41.7 31.7 47.4 51.5 18.7 51.3 18.4 28.2 6.6 49.3 36.7 5.2 45.2 58.9 6.6 11.7 52.4 23.5 66.8

116.965718 363.7030916 149.1255163 639.1437728 39.05977182 603.0076493 962.4804039 1243.330464 713.3875951 687.5470208 730.2992474 164.6569072 529.0322387 83.81335855 166.8561274 38.17496908 1707.75354 1189.826864 1749.23281

-5.86 -2.96 -0.44 0.04 0.47 -3.62 4.91 4.49 2.59 -5.1 5.05 0.81 -0.53 -5.72 2.36 2.8 1.52 2.25 -3.08

149000 275500 252700 150400 253800 236100 89900 509000 114900 148700 607700 158800 114500

17 37.6 40.7 21.3 48.9 39.4 14.7 58.1 21.3 29.8 54.8 22.7 17.7

136.6435059 281.5246485 441.328357 161.8328844 1101.429065 381.7203746 36.81241837 194.6532604 171.1141592 1708.09841 209.2477609 73.94159158 1320.371469

-1.3 -1.43 -1.3 -0.83 -2.15 -1.89 1.31 -0.37 1.73 0.75 -4.62 1.39 3.16

44

Hilton Head Island Hollywood Horry County Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia Brunswick Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

447900 200600 159600 792200 1000000 323300 360800 349200 167100 138300 248000 1000000 197100 109500 347500 705800 1000000 246100

48.5 28.4 22.7 65.8 80.4 60.6 32.2 59.5 27.3 19 31.9 69.1 34.6 10.1 38.7 68.7 75 31.9

896.9107079 203.7830872 237.4861052 931.7097566 147.9513863 223.2002234 67.31594177 1503.349026 1146.031891 1323.053682 669.4293444 146.8626322 562.3643219 194.9837121 316.2914699 286.6616795 716.9649675 1986.237614

-1.72 0.12 -0.9 -7.16 -6.45 0.42 1.37 -2.97 -3.77 -0.55 -2.85 -4.05 -3.94 0.39 0.18 -5.39 -7.17 -2.88

94700 154500 174500 155400 162900 125800 176100 145900 351700

12.3 19.5 31.3 16.8 26.1 22.6 39.4 26.1 37.6

900.3370068 82.37652825 617.9283562 109.3773321 189.6768047 1843.184584 645.6974236 1315.521404 1021.786465

2.97 -1.34 0.16 -0.66 0.29 -0.04 -1.4 0.59 -0.69

B.2 FLOODPLAIN COMMUNITY DATA

North Carolina Alliance Atlantic Beach Bayboro

FP Median FP Housing Educational FP Population FP Value Attainment Density SoVI 128800 14.6 371.974243 2.36 238733 25.72062084 556.6289014 -5.45 78100 6.1 681.7074889 2.36

45

Beaufort Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville Kill Devils Hill Kitty Hawk Manteo Minnesott Beach Morehead City Nags Head New Hanover County Newport North Topsail Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach

247240 86000 283300 277525 251758 423100 288800 166220 85700 245758 312292 573400 206050 466266 121088 469200 154300 107285 260075 332700 400000 176000 228300 262950 350230 230050

24.98063517 10.8 40.76923077 37.24241289 26.44636502 50.45731707 33.00153139 26.48597398 10.9 18.81460549 31.22674349 58.3 28.42424242 39.15391539 12.5 55.28455285 10.68775461 23.16021288 35.79608423 28.0390417 32.3 19.81799798 31.0802139 36.35676493 40.57978385 12.6142596

432.4881485 1060.697428 271.9748713 837.110163 56.85040984 175.0793714 389.1929822 136.7495952 487.6778419 62.18362035 46.55903261 152.7606505 259.1329631 500.7756712 115.5850911 272.6338658 6.81492619 746.5617972 844.8753807 250.3139602 757.5013891 61.40189916 631.0251775 110.0469974 465.1683359 400.2573489

-0.95 2.9 -1.53 -4.47 -0.43 -3.17 -1.61 0.53 2.58 -2.08 -2.57 -5.1 6.12 -4.69 -6.28 -3.02 1.04 -2.11 -4.67 -3.89 -1.63 0.32 -0.12 -1.87 -2.52 1.74

286800 248866 547433 322800 192100 393800 81400 210100 79200 461900 254866 235400 279550

41.7 39.6589659 47.36842105 51.5 23.75259875 51.3 23.29032258 34.49408673 8.268330733 48.66180049 31.38564274 19.3220339 23.3974359

116.965718 408.6095606 248.8096146 639.1437728 28.70489066 603.0076493 261.7946765 1496.213087 85.3061709 568.1002484 560.4272201 22.05247921 319.751808

-5.86 -1.92 -3.9 0.04 1.01 -3.62 4.91 4.49 2.58 -5.1 3.71 -0.24 0.09

46

Topsail Beach Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown Hilton Head Island Hollywood Horry County Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia

391500 75300 87966 236600 111420 881800

58.9 6.6 13.98392223 52.4 18.91288161 66.8360864

83.81335855 166.8561274 491.3425325 1707.75354 949.9070247 195.4315933

-5.27 2.36 3.75 1.52 2.93 -3.52

163200 297278 277175 205221 341042 336157 108650 542400 121770 192057 607700 266915 176666 503290 200600 176747 792200 1000000 323300 360800 398241 220422 116050 264300 1000000 203766 87600 347500 705800 1000000 291833

30.19145803 39.19625588 32.05972066 31.67394038 49.46492021 42.14602368 11.89631064 60.10230179 23.73154074 40.88014981 54.4 21.73811197 25.0363901 48.66669539 28.4 19.6254446 65.8 80.4 60.6 32.2 60.47800763 28.24911868 12.20435194 30.43606364 69.1 39.70873786 0 38.7 68.7 75 30.18469657

16.07785253 116.9538274 280.325594 89.60806965 935.5505273 106.7365807 16.55604054 113.9550223 630.7546871 843.2775245 209.2477609 35.57957432 832.7478719 699.435466 203.7830872 118.1835074 931.7097566 147.9513863 223.2002234 67.31594177 782.4496679 1139.701902 1140.7785 1022.254504 146.8626322 401.4278888 234.5432974 316.2914699 286.6616795 716.9649675 1552.821406

-1.92 -0.41 -4.56 -2.35 -2.79 -1.42 0.71 -0.37 1.19 -0.57 -4.62 1.3 3.07 -1.95 0.12 -0.63 -7.16 -6.45 0.42 1.37 -3.11 -3.87 -0.03 -0.03 -4.05 -0.97 -0.71 0.18 -5.39 -7.17 -2.47

47

Brunswick Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

105638 165736 248641 164209 203465 127683 160540 134850 366633

10.4534005 22.42713039 38.94578991 19.17926566 29.67148003 22.56086399 39.4 25.69787936 35.43478261

513.7495391 41.10875294 228.9979195 111.1566158 86.34539127 908.4705488 376.7628506 485.8635421 1051.932106

2.97 -1.37 -1 -0.59 0.15 0.14 -0.91 0.75 -0.69

B.3 300 SERIES CRS DATA – 2014 North Carolina

310 Elevat ion Cert

320 Map Info Service

330 Outrea ch Project s

340 Hazar d Disclo sure

350 Flood Prote ction Info

Alliance Atlantic Beach Bayboro Beaufort Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville

56 45 56 56 56 48 56 90 79 56 112 56 112 56 70 56 50 56 56 56 112

140 140 140 140 140 140 140 140 140 140 140 140 140 140 140 140 140 140 140 0 140

0 217 68 86 168 104 250 176 184 131 163 211 182 131 153 0 179 207 65 28 36

5 5 5 5 5 5 10 10 5 10 10 15 10 5 10 10 10 5 10 5 15

27 62 27 28 66 71 70 70 28 67 64 26 56 54 69 0 25 59 47 59 56

48

360 Flood Prote ction Assis t 0 0 0 0 0 0 0 0 35 0 0 48 0 0 0 0 0 0 0 0 0

300 Tota l

228 469 296 315 435 368 526 486 471 404 489 496 500 386 442 206 404 467 318 148 359

Kill Devils Hill Kitty Hawk Manteo Minnesott Beach Morehead City Nags Head New Hanover County Newport North Topsail Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach Topsail Beach Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County

56 56 56 56 56 56 70

140 140 140 140 140 140 140

212 264 213 0 203 246 105

15 5 5 5 10 10 10

84 76 69 27 80 77 53

59 0 0 0 0 0 0

566 541 483 228 489 529 378

56 56

140 140

182 207

10 10

32 20

0 0

420 433

56 56 56 56 56

140 140 140 140 140

78 129 68 68 144

10 10 5 5 5

59 73 27 27 78

0 62 0 0 0

343 470 296 296 423

56 74 56 56 56 56 56 56 56 56

140 140 140 140 140 140 140 140 140 140

211 177 211 180 102 68 0 167 68 211

15 5 15 10 10 5 10 0 5 15

26 49 26 27 31 27 41 20 27 71

48 0 48 0 0 0 0 67 0 48

496 445 496 413 339 296 247 450 296 541

56 54 56

140 140 140

105 74 187

5 5 15

55 51 73

49 59 0

410 383 471

95 102 80 56 46 70

140 140 140 140 140 140

285 167 173 0 297 296

5 10 66 10 5 5

102 83 92 27 66 95

63 0 0 0 45 68

690 502 551 233 599 674

70

140

154

5

58

0

427

49

Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown Hilton Head Island Hollywood Horry County Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia Brunswick Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

112 90 66 84 56

140 140 140 140 140

213 106 118 171 97

10 5 5 10 20

65 20 61 79 48

0 0 0 0 0

540 361 390 484 361

56 116

140 140

173 140

61 81

24 90

0 68

454 635

95 56 82 95

140 0 140 140

285 0 193 291

5 10 10 5

102 22 64 102

59 0 59 59

686 88 548 692

51 95 66 66 56 56

140 140 140 140 140 140

285 285 180 210 179 178

15 5 10 15 5 10

102 102 58 60 77 24

59 59 67 52 0 0

652 686 521 543 457 408

81 56 95 95 95 95 56

140 140 140 140 140 140 0

300 0 285 285 310 284 12

0 10 5 5 5 5 10

85 27 102 102 102 76 24

27 0 59 59 59 63 0

633 233 686 686 711 663 102

56 56 56 66

70 140 140 140

81 171 229 0

5 10 10 10

0 47 98 53

0 59 67 45

212 483 600 314

54 102 56 71 56

140 140 140 140 140

169 41 243 250 207

10 10 10 10 10

90 42 61 85 83

35 62 0 70 54

498 397 510 626 550

50

B.4 400 SERIES CRS DATA – 2014 North Carolina

410 Additiona l Flood Data 10 60 10 10 10 10 60 60 60 32 10 10 60 82 60 10 60 32 60 10 10 60 60 10 10 32 82 60

420 Open Space Preserv e 0 74 0 342 36 0 74 81 81 0 36 0 0 91 74 36 74 36 74 46 46 84 370 36 0 54 251 84

430 Higher Regu Standard s 224 284 224 101 151 206 258 186 409 99 271 278 257 237 333 239 548 95 236 122 360 387 345 90 224 171 385 453

440 Flood Data Mainte n 120 115 105 105 50 105 115 105 74 105 162 136 103 93 129 0 107 96 105 0 77 103 103 100 105 71 129 121

450 Storm water Manag em 55 30 55 30 30 30 65 30 30 75 30 30 30 55 45 30 30 75 0 30 30 60 60 30 55 75 80 30

Alliance Atlantic Beach Bayboro Beaufort Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville Kill Devils Hill Kitty Hawk Manteo Minnesott Beach Morehead City Nags Head New Hanover County Newport North Topsail

409 563 394 588 277 351 572 462 654 311 509 454 450 558 641 315 819 334 475 208 523 694 938 266 394 403 927 748

32 60

0 91

226 523

97 95

75 30

430 799

51

400 Total

Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach Topsail Beach Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown Hilton Head Island Hollywood Horry County

60 60 32 32 82 10 10 10 60 10 10 60 50 32 10

74 74 0 46 362 0 173 0 248 36 0 74 415 0 0

305 515 239 230 429 278 229 278 480 270 224 312 200 239 246

97 133 113 105 213 136 77 136 113 105 105 115 89 105 136

55 90 75 75 75 30 0 30 30 30 55 30 70 75 30

591 872 459 488 1161 454 489 454 931 451 394 591 824 451 422

10 10 60

36 124 291

131 131 295

118 100 103

75 75 55

370 440 804

10 10 10 0 10 115 10 10 10 10 10 0

61 80 44 46 36 46 305 0 0 454 83 46

351 490 231 375 169 914 232 406 45 224 467 202

120 113 113 120 96 133 69 161 55 105 90 113

60 307 220 80 40 370 40 30 40 75 75 75

602 1000 618 621 351 1578 656 607 150 868 725 436

0 35

36 389

48 287

60 162

70 225

214 1098

10 10

36 44

351 250

143 55

60 0

600 359

52

Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia Brunswick Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

10 10

106 86

150 302

80 118

20 60

366 576

10 10 10 35 10 10

36 145 44 251 36 46

312 426 588 654 330 264

130 139 77 194 113 113

60 60 75 240 92 225

548 780 794 1374 581 658

0 10 10 10 10 10 0

190 0 36 36 341 301 0

395 99 408 383 232 244 294

207 88 105 185 122 79 55

0 40 60 60 60 40 0

792 237 619 674 765 674 349

10 10 10 10 10 10 10 10 10

44 36 198 46 36 54 80 44 160

261 218 278 569 291 457 357 473 201

0 103 195 89 113 136 132 146 56

0 80 55 129 75 210 15 15 15

315 447 736 843 525 867 594 688 442

B.5 500 SERIES CRS DATA – 2014 North Carolina

510 FP Managem Planning

520 530 Acquisition Flood and Protection Relocation

540 500 Drainage Total System Maint

Alliance Atlantic Beach Bayboro Beaufort

0 136 0 0

0 0 0 0

15 30 30 30

0 0 0 0

53

15 166 30 30

Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville Kill Devils Hill Kitty Hawk Manteo Minnesott Beach Morehead City Nags Head New Hanover County Newport North Topsail Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach Topsail Beach

95 132 87 109 148 109 128 0 134 91 114 0 111 0 121 127 88 91 91 0 0 118 91 160 0 110

0 0 0 0 0 0 40 0 0 0 0 0 0 0 0 0 0 0 55 0 0 0 0 85 0 80

482 0 0 0 0 0 0 0 0 0 0 0 0 0 0 21 0 0 0 0 0 0 34 0 0 0

230 210 230 30 280 30 15 230 30 30 15 240 268 190 30 15 203 315 30 190 15 170 300 15 218 30

807 342 317 139 428 139 183 230 164 121 129 240 379 190 151 163 291 406 176 190 15 288 425 260 218 220

115 134 177 177 0 0 138 0 0 0 0 0 108

0 70 0 0 0 0 0 0 0 0 0 0 100

42 0 0 0 0 0 0 0 0 0 0 0 28

30 210 30 30 230 230 230 230 30 15 30 135 218

187 414 207 207 230 230 368 230 30 15 30 135 454

54

Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown Hilton Head Island Hollywood Horry County Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia Brunswick

177 0 113 113 110

0 0 0 180 70

0 0 0 59 234

30 230 280 230 30

207 230 393 582 444

260 0 0 0 260 310 103 101 0 0 260 151 0 170 260 105 260 260 260 260 260 80 260 252 151 0 260 260 260 260 105

0 0 20 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 25 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 84 0 0 0 0 0 0 155 0 0 0 0 0 0 0

315 300 268 15 268 270 253 180 0 243 30 0 265 255 315 0 315 280 315 315 203 147 210 280 280 0 315 315 330 268 0

575 300 288 15 528 580 356 298 0 243 290 151 265 425 575 105 659 540 575 575 463 252 470 687 431 0 575 575 590 528 105

0

0

0

0

0

55

Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

0 178 0 95 0 0 250 0

0 130 0 0 0 0 300 0

0 0 0 0 0 0 0 0

268 218 250 230 288 270 230 268

B.6 600 SERIES CRS DATA – 2014 North Carolina

610 Flood Warning

620 Levee Safety

630 Dam Safety

600 Total

Alliance Atlantic Beach Bayboro Beaufort Belhaven Cape Carteret Carolina Beach Carteret County Caswell Beach Cedar Point Craven County Creswell Currituck County Dare County Duck Edenton Emerald Isle Havelock Holden Beach Hyde County Jacksonville Kill Devils Hill Kitty Hawk Manteo

0 0 0 0 80 0 120 0 130 110 0 140 40 155 120 0 130 108 0 65 0 130 120 120

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

52 58 52 52 58 58 58 52 58 52 52 58 58 52 52 58 52 52 52 58 52 52 52 52

52 58 52 52 138 58 178 52 188 162 52 198 98 207 172 58 182 160 52 123 52 182 172 172

56

268 526 250 325 288 270 780 268

Minnesott Beach Morehead City Nags Head New Hanover County Newport North Topsail Beachtown Oak Island Ocean Isle Beach Oriental Pamlico County Pine Knoll Shores Plymouth River Bend Roper Southern Shores Southport Stonewall Sunset Beach Topsail Beach Vandemere Washington County Washington Park Washington Wrightsville Beach South Carolina Awendaw Beaufort County Beaufort Berkeley County Charleston Charleston County Colleton County Edisto Beach Florence County Florence Folly Beach Georgetown County Georgetown Hilton Head Island

0 105 120 0 0 40

0 0 0 0 0 0

52 52 52 52 52 52

52 157 172 52 52 92

40 110 0 0 145 131 0 140 120 0 0 0 130 0 111 0 60 130

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

52 58 52 52 52 58 52 58 52 52 52 52 58 52 58 52 52 58

92 168 52 52 197 189 52 198 172 52 52 52 188 52 169 52 112 188

130 67 60 0 180 205 75 125 0 0 0 0 0 100

0 0 0 0 0 0 0 0 0 0 0 0 0 0

40 40 40 40 40 40 40 40 40 40 40 40 57 40

170 107 100 40 220 245 115 165 40 40 40 40 57 140

57

Hollywood Horry County Isle of Palms Kiawah Islandtown Mcclellanville Meggett Mount Pleasant Myrtle Beach North Charleston North Myrtle Beach Pawleys Island Port Royal Ravenel Rockville Seabrook Island Sullivans Island Surfside Beach Georgia Brunswick Camden County Chatham County Effingham County Glynn County Hinesville Pooler Savannah Tybee Island

130 0 176 130 130 130 180 95 180 0 140 0 130 130 130 130 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

40 40 40 40 40 40 40 40 40 40 57 40 40 40 40 40 40

170 40 216 170 170 170 220 135 220 40 197 40 170 170 170 170 40

90 0 110 0 135 0 150 150 165

0 0 0 0 0 0 0 0 0

62 62 64 62 62 62 62 62 62

152 62 174 62 197 62 212 212 227

58

Loading...

Assessing Indicators of Community Participation, A Dasymetric and

University of South Carolina Scholar Commons Theses and Dissertations 5-2017 The Community Rating System: Assessing Indicators of Community Partici...

706KB Sizes 0 Downloads 0 Views

Recommend Documents

Community Health Indicators Toolkit - University of Regina
Cost of Living. Health Benefit Coverage. Local Availability. Cost of Food. Cost of Housing. Prescription Drugs. 1 – un

Indicators for assessing infant and young child feeding practices
Indicators for assessing infant and young child feeding practices part 2: measurement. 1.Infant nutrition. 2.Breast feed

Project Objectives, Indicators and Assessing Social - Better Evaluation
Working with primary stakeholders enables projects to be established in which the objectives are driven by local realiti

Adaptive Homestay as a form of community participation
pdf. Adaptive Homestay as a form of community participation & preservation Case Study: tourist village of Pentingsari in

The Construction of Community Participation - Semantic Scholar
KMNK/BKKBN (Kantor Menteri Negara Kependudukan/Badan Koordinasi Keluarga. Berencana Nasional) (1994b) Petunjuk tehnis id

Assessing the Impacts of Public Participation: Concepts - Ipea
DEFINE AND CATEGORIZE PUBLIC PARTICIPATION MECHANISMS MORE .... participation evaluation that includes context evaluatio

Dasymetric Stratification of a Flood Plain: Development and - MSpace
Jun 1, 2016 - Dasymetric Stratification of a Flood Plain: Development and Refinement of the HAZUS. Flood Mapping Tool fo

Dasymetric Mapping
relative magnitudes across the surface of the map. In contrast, the dasymetric method highlights areas of homogeneity an

Community participation as a strategy in reducing - BORA - UiB
transport systems, prenatal risk screening and maternity waiting homes. However, only ...... otda.depdagri.go.id/images/

Certificate of participation Certificate of participation
Certificate of participation. Child Protection Action Forum school. This is to certify that represented. AT THE YOUTH DI