University of Southern California Dissertations and Theses

Projecting vulnerability: a combined analysis of sea-level rise, hurricane inundation, and social vulnerability in Houston-Galveston, Texas

(USC Thesis Other)

Projecting vulnerability: a combined analysis of sea-level rise, hurricane inundation, and social vulnerability in Houston-Galveston, Texas

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content PROJECTING VULNERABILITY:
A COMBINED ANALYSIS OF SEA-LEVEL RISE, HURRICANE INUNDATION, AND
SOCIAL VULNERABILITY IN HOUSTON-GALVESTON, TEXAS
by
Susan Seymour
A Thesis Presented to the
FACULTY OF THE USC DORNSIFE COLLEGE OF LETTERS, ARTS AND SCIENCES
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE)
August 2023
Copyright 2023 Susan Seymour
ii

To my mom
iii
Acknowledgements
I am grateful for everyone who has supported me through my thesis. I would like to recognize
Professor Elisabeth Sedano for steering me in the right direction and assisting me when needed. I
would also like to thank the many other professors that made this experience worthwhile,
including Professor Andrew Marx and Professor Steven Fleming. A special thanks to Professor
An-Min Wu for a wonderful experience on Catalina Island and for starting this endeavor on the
right foot.
I am grateful to my husband, Tom Seymour, for enduring this arduous process, as well as
my friend Ramy Alyatim. A special thank you to my friend and coworker Jason Jordan, from the
Army Corps of Engineers, for help throughout the entire master’s program and especially during
my thesis. He has helped me tremendously including some python scripting, JavaScript, and
overall geoprocessing methods. I would also like thank Mike Koon from the Army Corps of
Engineers for helping me with geoprocessing and the python script that created the water surface
elevations. I could not have done this without them.

iv
Table of Contents
Acknowledgements ........................................................................................................................ iii
Table of Contents ........................................................................................................................... iv
List of Tables ................................................................................................................................. vi
List of Figures ............................................................................................................................... vii
Abbreviations ............................................................................................................................... viii
Abstract ........................................................................................................................................... x
Chapter 1 Introduction .................................................................................................................... 1
1.1 Motivation ........................................................................................................................... 2
1.2 Project Study Area .............................................................................................................. 3
1.2.1 Demographics ............................................................................................................ 4
1.2.2 Geophysical Attributes............................................................................................... 6
1.2.3 Project Overview ....................................................................................................... 7
1.2.4 Thesis Overview ........................................................................................................ 8
Chapter 2 Literature Review ......................................................................................................... 10
2.1 Sea Level Rise Projection ................................................................................................. 10
2.2 Mapping Inundation .......................................................................................................... 11
2.3 Cadastral-based Expert Dasymetric System (CEDS) ....................................................... 12
2.4 Socioeconomic and Locational Vulnerability ................................................................... 14
2.4.1 Modeling Social Vulnerability ................................................................................. 15
2.4.2 Socioeconomic Factors and Vulnerability Indexes.................................................. 16
2.4.3 Social Vulnerability Indexing and Weights ............................................................. 19
2.5 Population Growth and Exponential Smoothing Algorithm (ETS) .................................. 20
Chapter 3 Methods ........................................................................................................................ 22
3.1 Methods Overview ............................................................................................................ 22
3.2 Data ................................................................................................................................... 23
3.2.1 SLR and SLOSH ...................................................................................................... 24
3.2.2 Tax Parcel ................................................................................................................ 26
3.2.3 Census and Brown University Data ......................................................................... 28
3.3 Project Analysis ................................................................................................................ 29
3.3.1 Projecting Water Surface Elevations ....................................................................... 30
3.3.2 Mapping the Cadastral-based Expert Dasymetric System (CEDS) ......................... 33
3.3.3 Modeling Vulnerability ............................................................................................ 36
3.3.4 Population Growth and Exponential Smoothing Algorithm (ETS) ......................... 39
Chapter 4 Results .......................................................................................................................... 41
4.1 Water Surface Elevations .................................................................................................. 41
v
4.2 Cadastral-based Expert Dasymetric System ..................................................................... 45
4.2.1 CEDS Method Galveston County ............................................................................ 45
4.2.2 CEDS Method Harris County .................................................................................. 48
4.3 Vulnerability Index ........................................................................................................... 50
4.4 Population Growth and Exponential Smoothing Algorithm (ETS) .................................. 53
Chapter 5 Discussion and Conclusions ......................................................................................... 57
5.1 Study Findings .................................................................................................................. 57
5.2 Limitations and Considerations ........................................................................................ 59
5.3 Comparison Analysis and Conclusion .............................................................................. 61
References ..................................................................................................................................... 65
Appendices .................................................................................................................................... 70
Appendix A. Python Script ..................................................................................................... 70
Appendix B. Scale Factor Reclassification ............................................................................. 73

vi
List of Tables
Table 1. Explanatory variables ..................................................................................................... 18
Table 2. Explanatory variables reference table ............................................................................. 18
Table 3. Spatial and Tabular Data ................................................................................................. 24
Table 4. Harris County residential classification codes ................................................................ 27
Table 5. Vulnerability factors and indicator selection .................................................................. 37
Table 6. Harris County Exponential Smoothing Algorithm ......................................................... 54
Table 7. Galveston County Exponential Smoothing Algorithm ................................................... 55
Table 8. Repetitive Vulnerable Areas in Harris and Galveston County ....................................... 62

vii
List of Figures
Figure 1. Harris and Galveston, County ......................................................................................... 4
Figure 2. Harris and Galveston County, Texas flooding from Hurricane Harvey .......................... 7
Figure 3. Comparison of disaggregation methods. Maantay, Maroko, and Herrmann ................ 13
Figure 4. NOAA's operational storm surge basins ....................................................................... 26
Figure 5. Percent poverty in Harris County, Texas in 1960. Brown University and MapUSA ... 29
Figure 6. SLR and SLOSH method flow chart ............................................................................. 30
Figure 7. NOAA's SLR 2 feet ....................................................................................................... 30
Figure 8. SLOSH MOM Cat 5 High Tide..................................................................................... 32
Figure 9. Interpolated Water Surface Inundation - SLOSH and SLR two-feet ............................ 32
Figure 10. Census and parcel data preparation intersected with WSE flow chart ........................ 34
Figure 11. AHP indicator ranking scale example ......................................................................... 38
Figure 12. WSE at SLR zero feet for 2020 in Harris and Galveston County ............................... 42
Figure 13. Harris County WSE at SLR zero-, two-, three-, four, and five-feet ............................ 43
Figure 14. Galveston County WSE at SLR zero-, two-, three-, four, and five-feet...................... 44
Figure 15. Galveston County water surface population density map ........................................... 46
Figure 16. Galveston Island water surface population density map ............................................. 47
Figure 17. Harris County water surface population density ......................................................... 49
Figure 18. Analytic Hierarchical Process matrix results .............................................................. 51
Figure 19. Weighted overlay vulnerability indicator results ........................................................ 52
Figure 20. Density map and weighted overlay with a water surface elevation of five feet .......... 58
Figure 21. Geographically Significant Vulnerable Areas ............................................................. 63

viii
Abbreviations
3DEP 3D Elevation Program
ACS American community survey
AHP Analytic hierarchical process
ARA Adjusted residential area
ARIMA Autoregressive integrated moving average
CAT Category
CEDS Cadastral-based expert dasymetric system
ESL Extreme sea level
ETS Exponential smoothing
EWMA Exponentially weighted moving average
FEMA Federal Emergency Management Agency
GIS Geographic information science
IDW Inverse distance weighted
IPCC Intergovernmental Panel on Climate Change
MEOW Maximum envelope of high water
MHHW Mean higher high water
MOM Maximum of maximum
NTDE National Tidal Datum Epoch
NOAA National Oceanic and Atmospheric Administration
RA Residential area
RCP Representative concentration pathway
RU Residential units
ix
SLOSH Sea, lake, and overland surges from hurricanes
SLR Sea-level rise
TIN Triangulated irregular network
TNRIS Texas Natural Resources Information System
USC University of Southern California
USGS United States Geological Survey
WSE/I Water surface elevations/inundations

x
Abstract
Communities in the Houston-Galveston area of Texas are consistently at risk of hurricane
devastation. With warming climates and increasing greenhouse gases, sea-level rise (SLR) has
become a significant consideration. Many studies have shown the correlation between SLR and
vulnerability, however, little has been found on the implications of SLR with the influence of
storm surge on the community. This study established the current population and projected future
population at risk in 2050 and 2100 from SLR and storm surge inundation in Houston and
Galveston County. The National Oceanic and Atmospheric Administration’s (NOAA)
projections of SLR of two-, three-, four-, and five-feet are combined with NOAA’s Sea, Lake,
and Overland Surges (SLOSH) predictions to produce water surface elevations as sea level rises.
A social vulnerability index was created, and weights were determined, using an analytic
hierarchical process to reveal the socioeconomic vulnerable population within each water surface
elevation produced. A cadastral-based expert dasymetric system method was employed to
improve upon census data alone for spatial data of the population at 2020. An exponential
smoothing algorithm was then used to predict future populations utilizing census data from
Brown University and the American Community Survey from 1960 through 2020. The final
assessment establishes inhabitants who were at risk in 2020 and the projected population in 2050
and 2100 within rising sea-levels. The results identifies the neighborhoods within Harris and
Galveston County that are vulnerable to sea-level rise and storm surge inundation currently and
in the future. This provides these two counties, and other government agencies, a geospatial
assessment of vulnerable demographics within their locality and future estimates to assist in
planning, preparation, and emergency response.
1
Chapter 1 Introduction
The Texas Gulf Coast, specifically the Houston-Galveston area, has been impacted by climate
change and has repeatedly suffered immense storm surge inundations and flooding. Between
2015 and 2017, this area saw three 500-year flood events: the Memorial Day Floods, the Tax
Day Floods, and Hurricane Harvey (Boyer and Vardy 2010). The effects of sea-level rise (SLR)
will accentuate this risk since this is a coastal, low-lying area. As climate change has accelerated
over the last 20 years, the global mean sea level has risen exponentially ((National Oceanic and
Atmospheric Administration (NOAA) 2021). With these rising sea levels, this low-lying area
will observe persistent increases in flooding that will advance further inland. With the
Intergovernmental Panel on Climate Change (IPCC) projections of SLR increasing within the
next 30 to 70 years, future hurricanes and storm surge will likely devastate this coastal
community. (Oppenheimer et al. 2019)
Hurricanes Harvey, Ike, and Rita brought record breaking rainfall and significant storm
surge and they caused catastrophic flooding and billions of dollars in damage (NOAA National
Hurricane Center (NHC) 2021). Hurricanes such as these are projected to become more
common, devastating the surrounding areas and displacing millions of people (Carlson,
Goldman, and Dahl 2016). Socioeconomic hardship will accompany this flooding and the
population will become even more vulnerable. To identify the vulnerable people in this scenario,
it is important to recognize the correlation between geophysical and social systems.
(Chakraborty, Collins, and Grineski 2019). Both physical and demographic vulnerabilities are
essential in projecting the population at risk. With the combination of SLR, hurricane inundation,
and socioeconomic data, future storm surge inundation elevations and the vulnerable population
within these areas are measured in this project.
2
The analysis in this project used spatial and tabular data with Geographic Information
Science (GIS) tools to estimate future impacts of SLR on Houston and Galveston County. It
combined estimates of future SLR and hurricane storm surge to estimate the future areas subject
to high risk of flooding. The estimated current population and their socioeconomic status were
established using 2020 data. This demographic data was intersected with water surface
inundations to obtain the vulnerable population. Finally, the population within these projected
vulnerable locations was estimated for 2050 and 2100, for an overall assessment of the future
population at risk.
1.1 Motivation
The purpose of the project is to project the future populations that will be vulnerable to
SLR and storm-surge inundation in the Houston-Galveston area. Every year, between June and
November, hurricanes striking the gulf coast create a major issue and the associated risks are of
significant concern. The Galveston Hurricane of 1900, Hurricane Rita, Hurricane Ike, and more
recently, Hurricane Harvey, have devastated these communities. With warming global
temperatures and SLR, hurricanes are becoming more frequent. According to the IPCC, many
low-lying areas will experience rare Extreme Sea Level (ESL) events annually by 2050, like
today’s 100-year storm (Oppenheimer et al. 2019). By the end of the century, these storms will
be commonplace. Only a few studies about SLR are available beyond 2100, but it is likely that it
will continue to rise for thousands of years (Oppenheimer et al. 2019). The rate of loss of the
Antarctic Ice Sheet and the Greenland Ice Sheet renders uncertainty beyond 2100 (Oppenheimer
et al. 2019). As the ice sheets melt, sea levels rise, and hurricanes strengthen, storm surge will
increase and place more people in danger. Storm surge occurs when water rises above its typical
level, or astronomical tides, and spreads across land. As sea level rises, storm surge will intrude
3
even further inland. To mitigate the casualties and losses from SLR, it is imperative to be
proactive and to recognize what hazards exist and what actions can be taken.
Hurricane Harvey hit the Texas Coast in August of 2017 and was accompanied by
record-breaking rainfall and catastrophic flooding (Blake 2018). The damage and lives lost from
this were seen firsthand and the devastation left many in life-threatening conditions. Projecting
storm surge inundation with rising sea levels and defining the susceptible population will help
communities plan and prepare for the future and is a step towards social well-being.
1.2 Project Study Area
The study area for this assessment is the Houston-Galveston area of Texas located in
Harris and Galveston Counties as shown in Figure 1. Galveston is an island on the Gulf of
Mexico with a population slightly over 50,000 and is the main beach town for most of eastern
Texas. It sits between the Gulf of Mexico, West Bay, and Galveston Bay. It is separated by a
channel that leads to the Trinity Bay and the Houston Ship Canal, also known as Buffalo Bayou.
Houston lies northwest of Galveston and is the fourth largest city in the United States with a
population of approximately 2.4 million, and 4.7 million in the county (Population USA 2022).
The Houston-Galveston port district is the second largest port for the import and export of oil in
the country (US Energy Information Administration 2021). It also incorporates NASA’s Johnson
Space Center, is the second largest area for Fortune 1000 companies, and has the number one
cancer treatment center in the country (Visit Houston Texas 2021). Aside from economics and
infrastructure, the city has a very diverse demography with more than 145 languages spoken
(Visit Houston Texas 2021). The bustling economy and coastal amenities draw a diverse
population to the area.
4

Figure 1. The City of Houston, Harris County, and Galveston County
1.2.1 Demographics
When a hurricane strikes, the damages affect the inhabitants on an extensive scale by
destroying their economy, livelihoods and displacing countless families. Vulnerability,
specifically social vulnerability, is founded on the concept of resiliency and the capability to deal
with and recover from disasters and usually lies within impoverished communities that lack the
resources to prepare and respond appropriately. Rising sea levels, larger and more frequent
hurricanes, and the geophysical attributes of the locality, puts many in harm’s way.
Texas has seen a steady growth in population over many centuries and has a high influx
of immigrants which leads to diverse demographics. Yet, diverse demographics does not imply
lack of ability to prepare and recover from disasters. Harris County’s poverty rate is 8.6% and
5
their persons with disabilities and person 65 years and older are well below the national average
at 6.8% and 11.4%. The national average for persons with disabilities is 13.7% and 65 years and
older is 16% (US Census 2020). Galveston County has a similar demographic with the poverty
rate, persons with disabilities, and persons 65 years and older below the national average at
19.5%, 8.6%, and 15.2% (US Census 2020). With the above-mentioned demographics below the
national average, these communities seem capable of coping with flooding; however, this a
broad generalization of the study area. Neighborhoods in the east of Houston, like Port Houston,
East Houston, Downtown, and Fifth Ward are particularly vulnerable as referenced on
Disproportionately Impacted Communities – Houston Harris County Winter Storm Relief Fund.
These areas are where nonprofits state they would target services and is used to increase outreach
to areas in need (Houston Harris County Winter Storm Relief Fund 2023). According to data
from Brown University, described in more detail in Chapter 3 herein, large minority groups and
the elderly reside within comminutes in east, south, and southeast Houston.
Unlike Houston, Galveston’s diversity is more scattered throughout the county and not
constrained to specific neighborhoods, with a few exceptions. Low-income households are right
next door to early 1900’s remodeled estates. Section 8 apartments are being built within a mile or
two from the beach. According to a study from the Greater Houston Community Foundation in
partnership with Rice-Kinder Institute for Urban Research, Harris and Galveston County have an
SVI score of 0.72 and 0.58 (0 indicating the lowest vulnerability to 1 indicating the highest
vulnerability) (Understanding Houston 2023). This suggests that even though the statistics seem
to support resiliency, this area has a high degree of socioeconomic vulnerability.
6
1.2.2 Geophysical Attributes
Houston-Galveston is a low-lying coastal area with an average elevation of forty-nine
feet above sea level in Houston and seven in Galveston (US Climate Data 2021). It includes
numerous water bodies and expansive coastal marshes. Galveston Bay is the largest estuarine
system in Texas. It also receives runoff from the Trinity and San Jacinto Rivers (The Nature
Conservatory 2013). This leaves this area extremely susceptible to flooding and storm surge.
After the Galveston Hurricane of 1900, where more than 8,000 people died from storm
inundation, the City of Galveston constructed a 17-foot seawall to protect inhabitants from future
hurricanes (NOAA NHC 2021; Davis Jr. 1951).
Today there is a Corps of Engineers levee and two reservoirs, Addicks and Barker, to aid
in flooding and storm surge protection. Still, flooding has exceeded the expected floodplain
elevations numerous times. Due to heavy rainfall, Hurricane Harvey’s flooding went well
beyond Federal Emergency Management Agency’s (FEMA) depicted flood zones and
encompassed most of Harris and Galveston County as shown in Figure 2. According to Watson
(2018), Hurricane Harvey, a category (CAT) 4, produced the “largest rainfall recorded in
history” and hit the 500-year floodplain in some areas. The Saffir-Simpson Hurricane Wind
Scale rating for hurricane categories range between categories 1-5 depending on the strength of
the wind and the damage it can cause. At a CAT 4, a hurricane’s sustained wind speed is
between 130-156 miles per hour and will cause catastrophic damage (NOAA NWS). Hurricane
Ike, a CAT 2 that struck in 2008, caused 100-year storm surge levels and Hurricane Rita, a CAT
3 at landfall that struck in 2005, caused flooding of 10 to 15 ft above normal tide (Harris County
Flood Control 2021; NOAA NHC 2021). The low elevation of this area means even mild
hurricanes can wreak havoc.
7

Figure 2. Harris and Galveston County, Texas flooding from Hurricane Harvey. Federal
Emergency Management Administration

As a consequence of devastating floods that plague this coastal environment, land
subsidence has caused the coast to shift as much as ten feet in some areas (Galloway, Jones, and
Ingebritsen 1999). This region is also abundant with impervious surfaces and low infiltration
rates that makes flooding evident (Blessing, Sebastian, and Brody 2017). With rising sea levels
exceeding one inch per year, this area will continually be submerged. (Galloway, Jones, and
Ingebritsen 1999).
1.3 Project Overview
The analysis in this project uses spatial and tabular data with GIS tools and integrates
SLR, storm surge, and demographic variables to project future populations vulnerable to rising
8
sea levels in Houston and Galveston County, Texas. NOAA’s Sea, Lake, and Overland Surges
from Hurricanes (SLOSH) Maximum Envelope of High Water (MEOW) Maximum of MEOW
(MOM) Category 5 (CAT 5) high tide storm surge inundations model represents coastal storm
surge. NOAA’s SLR projected elevations of potential coastal inundation are combined with
SLOSH storm surge inundations to form overall projected water surface inundations. Different
SLR estimates were joined with SLOSH inundation to show impacted areas along the coast. The
water surface inundations were then intersected with block groups with accuracy improved using
the cadastral-based expert dasymetric system (CEDS) method. CEDS uses census blocks and tax
parcels to disaggregate data to the tax parcel level and reaggregate it back to the census block.
This process extracts residential lots and combines the census data to find a more accurate
representation of the population and their demographics. It is used herein to estimate the total
population within each SLOSH/SLR inundation for the year 2020. From that total population,
the communities most vulnerable to hurricane damage were estimated using a social
vulnerability index (SoVI). The SoVI factored in socioeconomic variables using data from the
US Census, American Community Survey (ACS), and Brown University. An analytic
hierarchical process (AHP) was applied to assess the relative importance of these variables. The
weights derived from the AHP were applied using a weighted overlay to find the most vulnerable
areas within the water surface elevation boundaries. An exponential smoothing algorithm (ETS)
was then applied to project the population within these boundaries for the years 2050 and 2100.
The collective elements examine the socioeconomic status of the total population within the
inundated areas at the current sea level and in rising levels for the years 2050 and 2100.
9
1.4 Thesis Overview
This thesis is divided into five chapters. Chapter 1 provided a description of the
motivation and background information for this project, the study area, and the attributes of the
Houston-Galveston area. Chapter 2 describes the related work and relevant details of SLR and
at-risk inhabitants. Additionally, it discusses previous works and the application of GIS and other
tools to obtain demographic data for projecting future vulnerable populations. Chapter 3 provides
the methodology of combining SLR and storm surge boundaries and estimating the
socioeconomically vulnerable peoples within these inundations. It provides the data used in the
analysis and the procedures for determining an accurate assessment of the current population and
the future population within these water surface elevations as sea level rises. Chapter 4 illustrates
the results for each of these scenarios. Chapter 5 discusses the overall results, limitations, and
other considerations. The final section compares the findings of this study with other studies
relevant to population vulnerability and flooding hazards within Houston and Galveston County.

10
Chapter 2 Literature Review
The Houston-Galveston area has struggled from the effects of hurricanes for centuries. This
coastal region has been studied extensively for flood risks, inundations, hurricane storm surge,
and SLR. Excessive damage from Hurricane Harvey in 2017 sparked investigations concerning
the effects of hurricane storm surge and flooding on local industry, natural environment, wildlife,
as well as social vulnerability and economic losses. Other studies focus on coastal vulnerability
from SLR and the imminent devastation from the superstorms it will produce. However, these
preceding studies have not combined SLR and storm surge inundation. SLR alone is a
considerable factor; however, storm surge plays an integral role in flooding and damages to the
local community. This literature review addresses topics including SLR, storm surge, social
vulnerability factors, the CEDS method, population growth, and estimating the future inhabitants
at risk within the water surface elevations.
2.1 Sea-Level Rise Projection
Sea-level rise (SLR) is a global phenomenon and a concern for all coastal communities.
According to the United Nations, as of 2017, over 2.4 billion people, or 40% of the world’s
population live within 60 miles of coastal area. (United Nations 2017). In the US alone, eighty-
seven million people or 29 percent, live in these areas. (US Census 2020) SLR is destroying
natural barriers, such as salt marshes, which not only protect the coast from natural disasters but
also sequester large amounts of atmospheric carbon dioxide (Conrad 2021). Changes in the
carbon cycle, along with methane and other gasses, is what many scientists claim is the major
cause of SLR. The effects, however, lead to severe storms, storm surges, flooding, and erosion.
This creates havoc not only on the environment but on the inhabitants of the coastal community.
11
According to the IPCC global mean sea-level rise (GMSL) is caused by the expansion of
ocean water and ocean mass gain. The major factors influencing this are seawater expansion
from temperature rise, known as thermal expansion, melting glaciers, and changing ocean basin
depths from Earth’s movement. (NASA 2021; NOAA 2021; Rahmstorf 2012). To project GMSL
rise, the IPCC assessment uses climate models with a variety of future scenarios for future
greenhouse gas emission rates, called representative concentration pathway (RCPs). It calculates
GMSL using different ranges of RCP’s; RCP 2.6 - RCP 8.5, RCP 2.6 being the lower rate of
greenhouse gas and the RCP 8.5 being the upper. The current projections estimate that sea-level
will likely rise between 0.24 m (0.79 ft.) and 0.32 m (1.05 ft.) by 2050 and 0.43 m (1.41 ft.) and
0.84 m (2.76 ft.) by 2100 (Oppenheimer et al. 2019). Horton (2020) claims, in a survey
conducted within the scientific community, the belief is that GMSL will likely rise higher than
the IPCC projects to between 0.63 (2.07 ft.) and 1.32m (4.33 ft.) by 2100. This project will use
the higher estimation of SLR utilizing two- and three-feet for the year 2050, and four- and five-
feet for 2100 with the National Oceanic and Atmospheric Administration’s SLR layers to
conduct its study.
2.2 Mapping Inundation
Many analyses rely on FEMA’s floodplain maps to depict flood extent and potential risk.
However, according to many studies these maps are inaccurate. The 100-year floodplain is an
inadequate predictor, and a great deal of flooding happens outside the FEMA zones (Blessing,
Sebastian, and Brody 2017). FEMA’s flood maps indicate that approximately 15 million people
live within the 100-year flood zone. However, Smiley (2020) states that he believes this is
inaccurate and that other studies found this number to be around 1.7-3.1 times higher. New
models are being developed and indicate that twice as many properties are damaged from flood
12
inundation and approximately 47% of claims made to FEMA were outside the zone (Smiley
2020). Other studies have incorporated the Hydrologic Modeling System (HEC-HMS) and the
River Analysis System (HEC-RAS) developed by the US Army Corps of Engineers (HEC) to
include factors not considered in FEMA’s assessment (Blessing, Sebastian, and Brody 2017;
Bass and Bedient 2018). Each model provides different approaches to account for flood and
storm surge hazards; however, to apply localized data specific to the Gulf Coast, NOAA’s
SLOSH MOM High Tide Cat 5 for the Texas Basin was used.
SLOSH models are simulations of hurricane surges developed by the National Weather
Service using a multitude of factors developed for specific areas. These models include elements,
such as, tide levels, forward speed, storm categories, atmospheric pressure, and more localized
data, like levees, rivers, bridges, etc. (NOAA SLOSH). Maloney and Preston (2014) used
NOAA’s SLOSH data to estimate storm surge and SLR vulnerability along the Atlantic and the
Gulf Coasts following NOAA’s guideline, Mapping Coastal Inundation Primer (NOAA 2012).
NOAA’s guideline examines different approaches in creating inundations using their simulated
SLOSH data. NOAA’s approach for modeling water surfaces was used in this study with the
additional steps of incorporating SLR into the SLOSH layer. SLOSH High Tide Cat 5 was used
to show the worst-case scenario for the region.
2.3 Cadastral-Based Expert Dasymetric System (CEDS)
The CEDS method disaggregates data to a smaller unit of measure to obtain a more
precise understanding of population and US Census Bureau data. According to Maantay,
Maroko, and Herrmann (2007) this method differs from other dasymetric mapping techniques in
that it is more detailed and is particularly useful in “estimating population distribution in hyper-
heterogeneous urban areas” (Maantay, Maroko, and Herrmann, 2007, 85). Their study on
13
mapping population distribution in the urban environment shows how the CEDS method is more
beneficial than other methods in estimating population because it uses detailed cadastral data.
The CEDS uses tax lot data and residential units (RU) to analyze population. Including RU in the
analysis ensures the full population is accounted for in each tax lot by incorporating the
inhabitants and not simply the distribution within US Census Bureau blocks or tracts. The
Maantay, Maroko, and Herrmann (2007) study incorporated a buffer around high air pollution
areas and compared the methodological differences between Aerial Weighting, Filtered Aerial
Weighting and CEDS. Figure 3 shows their results and the benefits of using the CEDS method.

Figure 3. Comparison of disaggregation methods. Maantay, Maroko, and Herrmann

Miyake et al. (2010) uses the CEDS method in multiple studies to analyze the
demographic composition of populations within specific areas. The one disadvantage of this
method is that by estimating the population using residential tax lots within the same US Census
14
Bureau block group does not make the sub-populations independent from one another (Miyake et
al. 2010). Therefore, the data needs to be reaggregated back to the level it began at. Another
issue when assessing population and demographics is that not all populations are represented.
Typically, the poor, homeless, undocumented immigrants, and other marginalized peoples are
unaccounted for (Miyake et al. 2010).
Maantay, Maroko, and Herrmann’s (2007) study compared Filtered Areal Weighting
(Binary Method), adjusted residential area (ARA), RU and dasymetric mapping and found CEDS
to be the most accurate and best method for population distribution. Zoning, land use, lot size,
and RU were used in their study. This study uses the CEDS method starting with the US Census
Bureau blocks and parcels from both Harris and Galveston County. It does not use zoning since
Harris County does not have zoning regulations.
2.4 Socioeconomic and Locational Vulnerability
Coastal communities, like the Houston-Galveston area, are prone to hurricane flooding
and storm surge. This suggests susceptibility based on physical location, or what Logan and Xu
(2015) refer to as locational vulnerability. Vulnerability comprises many factors, and according
to most scientists it is defined as conditions or exposure to hazards and the sensitivity and the
resilience to it. (Turner 2003; Kasperson, Kasperson, and Turner 1995; Cutter, Mitchell, and
Scott 2000; Yuan, Guo, and Zhao 2017) To accurately determine the extent of loss, both
locational and socioeconomic vulnerability need to be considered.
The geophysical and socioeconomic environment are interdependent. Social vulnerability
depends on the capability of the community or individuals to adapt to the environment.
According to White and Hass (1975), population shifts, increased mobility, industrialization,
economic factors, and housing increases and needs, are the basis of the nation’s vulnerability to
15
hazards (as cited in Cutter, Mitchell, and Scott 2000, 714). Shifts in evolution forces society to
make changes that may not be desirable to accommodate ones’ needs. For example, moving to a
hazardous coastal community for employment and economic purposes. These socioeconomic
aspects are intertwined with locational vulnerability, or geographic vulnerability, to create what
Cutter, Mitchell, and Scott (2000) calls the overall place vulnerability. This project assesses the
overall place vulnerability of the communities within Harris and Galveston County. It signifies
which inhabitants are currently exposed and susceptible to storm surge and flooding and their
socioeconomic status. The population in 2020 is represented and the estimated population in
2050 and 2100 to illustrate continued SLR and determine overall place vulnerability.
2.4.1 Modeling Social Vulnerability
Social vulnerability modeling is a difficult task as there is not one set of indicators to
assess this. The Houston-Galveston area and its susceptibility and resilience to hurricanes defines
the social conditions of this community. In other words, the socioeconomic influences affect the
populations’ ability to cope with or recover from these disasters. This can include financial
hardship, disabilities, or education. Chakraborty, Collins, and Grineski (2019) assessed the
environmental justice implications of Hurricane Harvey flooding and find that Black and
Hispanic populations and socioeconomically deprived neighborhoods were the most vulnerable
and received the most flooding, where the more affluent have the means to move away from
these hazardous environments. (Chakraborty, Collins, and Grineski, 2019) However, Cutter,
Mitchell, and Scott (2000) state that in South Carolina the mean housing value is highest near the
coast where predominantly White populations are found. The socioeconomic determinants in
each of these instances are showing different outcomes for different areas. How these factors are
decided are influenced by the various attributes being measured.
16
In the scientific community, social vulnerability is most often determined using a SoVI.
Cutter, Boruff, and Shirley (2003) adapted a SoVI to consider other vulnerabilities besides
biophysical in a study on environmental hazards in the US Others have followed suit and both
Burton (2010) and Flanagan et al. (2011) used a SoVI to study the impacts on the community
from Hurricane Katrina in 2005 along the Mississippi’s coast to help aid governments agencies
and emergency management. To identify the location of socially vulnerable peoples in Harris
and Galveston County a SoVI was designed to measure demographic characteristics and explore
the population within water surface inundations with rising sea levels.
2.4.2 Socioeconomic Variables and Vulnerability Indices
To determine which socioeconomic variables are significant within the Houston-
Galveston area, the four stages of the disaster cycle used by emergency management personnel
to establish risk, are considered. These are Mitigation, Preparedness, Recovery, and Response. A
community that can withstand the consequences of a disaster in all four categories are more
resilient, while the inhabitants that lack these abilities are more susceptible to devastation. A
SoVI helps determine the population that is more susceptible. According to Flanagan et al.
(2011), a SoVI consists of four categories or domains that portray the major subcomponents of
establishing risk for disaster management. They are socioeconomic status, household
composition and disability, minority status and language, and housing and transportation.
Socioeconomic status includes factors like income, poverty, age, education, disability, and
employment. Low-income households may not have transportation or the ability to evacuate.
Poverty limits resources and can create homelessness, food shortages, health issues, and the
inability to seek aid (Flanagan et al. 2011). The elderly, young, and disabled are at a
17
disadvantage and may need support during disasters, such as medical care or transportation.
These disadvantages lead to the inability to prepare and recover from disasters.
The four domains listed above are the basis for constructing a SoVI and generating
explanatory variables within this project as shown in Table 1. A study conducted by
Chakraborty, Collins, and Grineski (2019) on the implications of Hurricane Harvey flooding on
the Greater Houston Area used five explanatory variables to create an index of significant
socioeconomic factors, no high school education, limited English language proficiency, income
below poverty level, no vehicles, and unemployment. They found these variables to be
significantly associated with their flood extent and comprised a majority of the population within
these neighborhoods. Flanagan et al. (2011) used 15 explanatory variables to create a SoVI for
disaster management with a case study on the impact of Hurricane Katrina. This study focused
on deaths related to drownings and found that the elderly was the most impacted. Most residents
were in nursing homes which correlates with the inability to evacuate without support. Table 2
references each vulnerability variable chosen and the study to which it relates. In each instance,
once the explanatory variables are decided, an index was created, weights were assigned to
vulnerability indicators, and a percentile rank was established.
18
Table 1. Explanatory variables
Groups Variables REFID
Socioeconomic status
below poverty level/low income 1, 2, 3, 5, 6
unemployed 1, 3, 4, 5
no high school 1, 2, 3, 4, 5
Household composition and
disability

elderly (65 and over) 1, 2, 3, 5, 6
younger than 5 2, 3, 5, 6
disabled 1, 2, 5
single parents 1, 5
renting 3, 5, 6
persons in group quarters 1
Minority status and language
Black/African American 1, 2, 3, 4, 5
Asian 1, 2, 3, 4, 5
Hispanic 1, 2, 3, 4, 5
do not speak English well/at all 1, 2, 4, 5
female 3, 5, 6
Housing and transportation
no vehicle 1, 3, 4, 6
proximity to pub transportation/number of bus
stops

Table 2. Explanatory variables reference table
REFID Source Article
1 Flanagan, et. al. A Social Vulnerability Index for Disaster Management
2
Bodenreider, et.
al
Assessment of Social, Economic, and Geographic Vulnerability Pre- and
Post-Hurricane Harvey in Houston, Texas
3
Fucile-Sanche,
Davlasheridze
Adjustments of Socially Vulnerable Populations in Galveston County,
Texas USA Following Hurricane Ike
4
Chakroborty, et.
al.
Exploring the Environmental Justice Implications of Hurricane Harvey
Flooding in Greater Houston, Texas
5 Cutter, et.al. Social Vulnerability to Environmental Hazards Index
6 Li and Lam
A spatial dynamic model of population change in a vulnerable coastal
environment
19
2.4.3 Social Vulnerability Indexing and Weights
The three most common approaches when creating an index are deductive, hierarchical,
and inductive (Tate 2012). The deductive approach was typically applied in earlier SoVI indexes
and usually contains ten or less indicators (Cutter, Mitchell, and Scott 2000; Montz and Evans
2001; Wu et al. 2002; Dwyer et al. 2004; Collins et al. 2009; Lein and Abel 2010, as cited in
Tate 2012). This approach uses variables from accepted universal knowledge. The hierarchical
method typically consists of ten to twenty indicators and can contain sub-indices within the
index. The inductive approach consists of twenty or more indicators and is the basis for Cutter’s
SoVI index that has been used in numerous studies. The hierarchical method has proved to be an
effective method for decision making and prioritizing by pairing indicators. This was the method
used in this study.
In a multicriteria analysis, such as this one, the AHP is a technique to quantify the
weights of each indicator against each other and determine the relative importance of the
relationship. It correlates each indicator and the weight assignments through a comparative
matrix. The resulting weights of the AHP are based on a pairwise comparison of the criteria and
a principal eigenvector value of greater than one to indicate independent indicators.
Applying weights to explanatory variables helps determine the importance of each
indicator. The more important the variable, the heavier the assigned weight. There is no
recognized methodology on how to construct an index; however, past studies have introduced
some criteria for ranking each variable (Tate, 2012; Cutter, 2000; Yuan, 2017). Using judgment
to assign relative importance is subjective; however, according to Tate (2013) it is comparable to
assigning equal weights to each indicator. Gathongo and Tran (2020) used the AHP method in a
study to assess social vulnerability in Kenya by assigning weights to the exposure, sensitivity,
and adaptive capacity of villages. They followed Saaty’s (2008) weighting method (1-9) to
20
assign weights by level of hierarchy; the hierarchical method. By obtaining a consistency ratio
under 10%, Gathongo and Tran (2020) surmised their judgment of selected indicators to be
satisfactory. The benefit of weighting using an AHP is that it quantifies subjective data using a
statistical process to recognize the relative importance of each indicator. The output of the AHP
assigns each weight a percentile rank to create an index ranking indicator set, i.e. a SoVI.
2.5 Population Growth and Exponential Smoothing Algorithm (ETS)
Population growth fluctuates and is dependent on many factors, birth rate, death rate, rate
of immigration, ecological systems, environment, economy, food supply. Different formulas
have been used to project future population; Percent Change, Linear Growth, Arithmetical
Increase or Arithmetical Mean Method, but the most common methods area the Autoregressive
Integrated Moving Average (ARIMA) and the ETS. According to Winters (1960), the
exponential smoothing forecasting model or ETS has advantages over conventional models. It
has better results, requires less information, and responds faster to shifts in the time series. It is
also non-stationary as compared to the stationary ARIMA model.
The ETS method originated with a US Navy analyst Robert G. Brown during World War
II (Bayak 2022; Gass and Harris 2000; as cited in Gardner 2006). He developed a method to
incorporate trends and seasonality into the ETS equation. Holt continued work on the ETS
method and developed his own version for dealing with seasonal data. Winters tested Holt’s
work and this method became known as the Holt-Winters forecasting system (Gardner 2006).
This model forecasts time series by utilizing three attributes: “a typical value (average), a slope
(trend) over time, and a cyclical repeating pattern (seasonality),” known as the Triple
Exponential Smoothing Formula (SolarWinds 2019). An Exponentially Weighted Moving
Average (EWMA) applies weights to values or attributes to smooth a time series. It weighs
21
recent data more heavily than older data. The Triple Exponential Smoothing Formula applies the
EWMA for each of the three attributes, average, trend, and seasonality (SolarWinds 2019).
Since the origination of ETS, it has become the prominent method in doctoral programs,
business forecasting, planning and budgeting, traffic-flow forecasting and many other time
series-based approaches. It has been incorporated into numerous programs and software and
Microsoft Excel has a function which runs an ETS (FORESCASTS.ETS). Baykal, Colak, and
Kılınc (2022) used this technique to forecast climate boundary maps from 2021-2060 as it
accounts for the alpha, beta, and gamma, or triple AAA values, and minimizes the mean global
error. Excel uses the target date (value to predict, date/time or numeric), value (historical values),
timeline (range of numeric values), seasonality (length of the season), data completion (accounts
for missing data values), and aggregation (aggregates multiple points with the same time stamp)
to calculate the forecast (Microsoft 2021). Utilizing the ETS in this study accounts for the
projected population distribution within each SLR rise elevation for the year 2050 and 2100.

22
Chapter 3 Methods
The goal of this project is to identify localities of people vulnerable to hurricane storm surge and
SLR in Harris and Galveston County. Water surface elevations were identified for five SLR
scenarios. The current at-risk population was ascertained for 2020 and the future population
within these inundations were determined. The methodologies for each process are described in
this chapter beginning with an overview of the project and the data used. The project analysis
section describes the four analyses applied to obtain the final results; the generation of the water
surface elevations, the CEDS method, the creation of the SoVI, the future population growth
determined by an ETS.
3.1 Methods Overview
This analysis began with the creation of water surfaces from hurricane storm surge and
SLR in Harris and Galveston County. It uses NOAA’s SLOSH MOM Cat 5 High Tide storm
surge inundations as a baseline, subsequently referred to as SLOSH. Current sea level is
represented as SLR zero feet, while two- and three-feet SLR layers are used for 2050 and four-
and five-feet for 2100. Each SLR layer is combined with SLOSH storm surge showing the
respective scenarios of inundations. Current SLR at zero feet is combined with SLOSH
inundations and are compared to two- and three-feet SLR layers for 2050 and four- and five-feet
for 2100.
A CEDS method was employed for a more accurate estimate of total population in 2020
by disaggregating the data to the tax lot level and reaggregating it back to the census block
group. This intersection of this data with the WSE represents the current inhabitants affected. To
further ascertain the populaces at risk, a SoVI was created. This established the demographic
23
indicators and the AHP method was then employed to determine percentile ranks for each.
Sixteen variables are explored, and their importance weighted, and a weighted overlay illustrates
the most vulnerable areas within each inundation level. This data describes the populace, and
their social standing, which resides within the potential risk area.
Brown University data, containing ACS data from the years 1960 through 2010, and US
Census data for 2020 was used to project future populations utilizing an ETS. The ETS leverages
past population data to project the future population. The final assessment represents the current
vulnerable populace in 2020 and the projected vulnerable population in 2050 and 2100 within the
estimated sea level rise elevations.
3.2 Data
The data for this project consisted of tabular and spatial data in both vector and raster
format. The data names, types, scale, coordinate system and source are listed in Table 1. The
spatial data comprised of sea level and elevation data, census block boundaries, tax and land use
parcels, and bus stop locations that were included in the SoVI. The tabular data used was US
Census, ACS, and Brown University Data census data.
24
Table 3. Spatial and Tabular Data
Data Type Scale
Original Coordinate
System Source

Date
SLOSH MOM
High Tide Cat 5 Raster
Atlantic &
Gulf Coast NAD 1983 NOAA

2012
SLR Raster Multiple NAD 1983 NOAA 2016
DEM Raster
Northeast
Texas NAD 1983 USGS

2018
Census Tabular
Block
Group &
Tract -
US Census
Bureau

2020
ACS Tabular
Block
Group &
Tract - ACS

2020
Brown University
(1950-2010) Tabular Tract level - NHGIS
1950-
2010
Tiger/line
shapefiles Vector
Block
Group &
Tract NAD 1983
US Census
Bureau

2020
Brown University
Tracts (1950-2010) Vector Tract level NAD 1983 NHGIS

1950-
2010
Harris and
Galveston County
Boundaries Vector County WGS 1984
Harris Central
Appraisal
District

2020
Harris County tax
lots Vector Parcel
NAD 1983 StatePlane
Texas S Central FIPS
4204 (US Feet) TNRIS

2020
Galveston County
tax lots Vector Parcel
NAD 1983 StatePlane
Texas S Central FIPS
4204 (US Feet)
Galveston
Central
Appraisal
District

2020

3.2.1 SLR and SLOSH
SLR inundations from NOAA are in one-foot increments from 1-10 feet of rise
inundation extent. NOAA’s SLR depth grid raster shows inundation extents at the current mean
higher high water (MHHW) level, or the mean of the higher tidal water heights over the National
Tidal Datum Epoch (NTDE) in a tidal day (NOAA Tides and Currents). NOAA creates the tidal
model, using their VDATUM transformation software, to represent the MHHW in orthometric
25
values or North American Vertical Datum of 1988 (NOAA Tides and Currents). This data
illustrates the potential flooding within certain coastal areas. The SLR layers for Brazoria,
Chambers, Galveston, Harris, and Liberty Counties were downloaded from NOAA’s Office for
Coastal Management Sea Level Rise Data 1-10 ft Sea Level Rise Inundation Extent, located on
their InPort website hosted by NOAA Fisheries. (NOAA Office for Coastal Management). The
SLR elevations of zero, two-, three-, four- and five-feet were chosen to reflect the IPCC and
other estimates within the scientific community’s assessment of projected SLR in 2050 and
2100.
NOAA’s SLOSH MOM data represents hypothetical storm surge extents using a
computerized model to analyze elements like atmospheric pressure, forward speed, and historical
track data (NOAA SLOSH). The SLOSH layer depicts the worst-case scenario from high water
values to show flooding at certain locations. The available basins coverage from NOAA is
shown in Figure 45. The SLOSH MOM Category 5 High Tide is used in this study to depict
worst-case scenario inundation levels.
26

Figure 4. NOAA's operational storm surge basins. Source: NHC

3.2.2 Digital Elevation Model
A United States Geological Survey (USGS) 1/3 arcsec (10m) DEM was downloaded
from the USGS National Map website. The DEM is a product of the 3D Elevation Program
(3DEP) managed by the USGS providing high-quality lidar elevation products nationwide. The
10m DEM has the most coverage and is the highest resolution seamless DEM provided by the
3DEP service (USGS 3DEP). This DEM was used in the inundation analysis process to subtract
land values from the SLR and SLOSH combined layers to produce a final water surface
elevation.
27
3.2.3 Tax Parcel Data
Tax parcel data was obtained for the CEDS method, and the residential lots extracted to
represent the population. The parcel data for Harris County was obtained through Harris Central
Appraisal District and supplemented with Texas Natural Resources Information System (TNRIS)
data. The TNRIS data contained the land use codes and was needed to determine residential lots.
Harris county classifies residential lots into six categories as shown in Table 2. All the residential
categories were used in this study to fully represent the population within the county regardless
of single or multifamily units for a total of 721,253.
Table 4. Harris County residential classification codes
A1 Single-Family
A2 Mobile Homes
B1 Multi-Family
B2 Two-Family
B3 Three-Family
B4 Four- or More-Family

Galveston County data was obtained through the Galveston Central Appraisal District.
Their land use categories for residential classifications only consist of one, RL for residential lot.
Galveston County had a total of 121,531 residential lots.

3.2.4 Census and Brown University Data
Census data was from the US Census Bureau, the ACS, and Brown University (credited
to the National Historical GIS). The census data provided demographic data in 2020 for use in
28
the CEDS method. To follow the CEDS method the data is disaggregated from the block group
level to the tax lot or parcel level and was therefore downloaded as block groups.
The ACS, established by the US Census, provided data in five-year estimates and was
used for supplemental data when needed. These data were also obtained at the block group level.
The Tiger/Line shapefiles, also a subset of the US Census, was downloaded and joined to that
tabular data to provide a geographical reference to the demographic data.
The Brown University demographic data was obtained from a MapUSA project on
diversity and disparity (IPUMS USA). This dataset is credited to the US Census Bureau. The
project, called A Human Mapping Project (1940-2010) entailed demographic data, to include
Harris and Galveston County, from 1940-2010. The census data for these years are maintained
by the National Archives and Records Administration but have limited accessibility and
demographic data (Census History). The Human Mapping Project contained the demographic
data needed for this project along with the geospatial data for the coinciding year. Information
about this project can be found through Brown University and MapUSA. Figure 5. shows a
section of Harris County, Texas depicting the percent in poverty in 1960 created from the Brown
University data.
29

Figure 5. Percent poverty in Harris County, Texas in 1960. Sources: Brown University and
MapUSA

3.3 Project Analysis
This section describes the tabular and GIS data integrated to create the water surface
elevations from SLR and SLOSH data. It then discusses the CEDS method employed to obtain a
more accurate assessment of the current population. Additionally, the SoVI, AHP, and weighted
overlay analysis portrayed the at-risk population and finally, the ETS and population growth is
explored.
30
3.3.1 Projecting Water Surface Elevations
This analysis combines NOAA’s zero-feet SLR for current conditions, two- and three-
feet for the year 2050, and four- and five-feet for 2100, with SLOSH data to conduct its study.
The flow chart in Figure 5 shows the methodology used to create water surface inundations for
each sea level rise instance. The SLR layers are prepared using ArcGIS’s SetNull tool
1
to remove
invalid or no data values. The same tool is also run on the USGS DEM to remove null values.
Both rasters are reprojected into NAD1983 State Plane Texas South Central FIPS 4204 Feet to
match the Harris and Galveston County data. The Times tool is then used to convert elevations
from meters to feet by multiplying by 3.28083333.

Figure 6. Workflow for projecting SLR and SLOSH inundation boundaries

1
All tools refered to hereafter are ArcGIS tools.
31
The SLOSH data preparation consisted of converting the hightide SLOSH MOM grids to
points. Elevations were obtained from the original SLOSH MOM grid. The point layer was then
reprojected into NAD1983 State Plane Texas South Central FIPS 4204 Feet and a new field is
added to multiply the values by 3.28083333, from meters to feet. To create a smooth raster
surface by interpolating the extracted points, a second order inverse-distance weighted (IDW)
was applied. Given the density of the points and known z values, this method is the most
appropriate to interpolate this data. Prior to running the IDW a triangulated irregular network
(TIN) dataset was created by importing the points. A TIN domain was then generated to create a
polygon that represents the interpolation area. The DelineateTinArea tool was used to create a
polygon around the perimeter of the TIN or interpolated point area. This allows the IDW to
interpolate the area appropriately and not connect unrelated points. The polygon was then
converted to polylines. Before running the IDW tool, random points were selected and removed
from the point layer to use as checkpoints to evaluate the final IDW layer elevations. A second
order IDW was run with the TIN domain polyline as the input barrier. The output created a
smooth water surface elevation from the original SLOSH grid that indicates hurricane storm
surge from a CAT 5 at high tide. This is then merged with each SLR elevation and demonstrates
how storm surge is intensified with the inclusion of SLR.
The SLOSH and SLR data were combined with the Mosaic to New Raster tool, using a
mosaic operator of sum and a processing extent of Union of Inputs. This created one raster with
the sum of elevations and extent of both the SLOSH and SLR layers. The Raster Calculator was
used to subtract the DEM from the new combined raster resulting in a surface inundation
(NOAA 2012). The SetNull tool was applied to remove values that did not represent water
inundations and the final output is an interpolated water surface. This process is repeated for
32
each SLR increase producing a total of five interpolated water surfaces. Figure 6 demonstrates
SLR at an elevation of two feet, Figure 7 is the SLOSH MOM layer Cat 5 High Tide, and Figure
8 represents the interpolated water surface, water surface elevation, of the combined SLR at two
feet and SLOSH layer.

Figure 7. NOAA's SLR 2 feet Figure 8. SLOSH MOM Cat 5 High Tide

Figure 9. Interpolated Water Surface Inundation - SLOSH and SLR two-feet
33

The process of creating an interpolated surface using IDW, combining the SLOSH and
SLR data, subtracting the DEM, and setting null to non-water surface values was done using
ArcGIS Pro Python 3 (Arcpy) as shown in Appendix A.
3.3.2 Mapping the Cadastral-based Expert Dasymetric System (CEDS)
The CEDS methods uses 2020 census data and disaggregates the data to the tax lot level
for a more accurate assessment of population and its attributes. Data preparation for census data
consisted of joining the geospatial block groups with the tabular data using the “GEOID” Codes.
The Harris County parcel data is joined with the TNRIS parcel data to add the land use codes to
determine residential lots. The land use code field for Harris County is the “state_land” and
Galveston is the “landuse.” The population census data is then spatially joined with the parcel
data using the Intersect tool and clipped to the county boundary. This combined data creates a
new layer for each county, one for Harris and one for Galveston. Residential lots are extracted
from each county to account for only residential land use. These new layers are intersected again
with the interpolated water surface inundations for each rate of emissions, creating ten new
layers, five for each county. Figure 10 is a diagram of the process taken to prepare and combine
census and parcel data and then the intersection of the results with the water surface inundations.
34

Figure 10. Census and parcel data preparation intersected with WSE flow chart

Once the residential parcels are intersected with the water surface inundations, statistics
are calculated for the RU and residential area (RA), using summary statistics, and a sum value is
returned for the RU and RA fields. If the parcel RA value is null or zero but census data is
showing RU, then an ARA is necessary to account for missing data. The equation for calculating
the ARA is:
ARA = M * (BA * RU / TU) + RA
IF RA = 0 and RU <> 0, THEN M = 1, ELSE M = 0

where BA is the building area, TU is the total number of units, and M is binary value designation
ancillary data for ARA. (Maantay, Maroko, and Herrmann 2007). If the calculated difference
between each estimated population is less than or equal to the ARA, then the RU population is
used. Otherwise, the ARA value is used as the “superior proxy unit” (Maantay, Maroko, and
35
Herrmann 2007, 88). Once it is established if the RU or ARA will be used the derived population
needs to be calculated. This project uses the RU to determine population as there were no
missing or null values within the residential data.
To find the RUs, the Dissolve tool is used. This tool obtains the Ul and Uc values for each
of the ten new layers, five in Harris County and Five in Galveston County, to find the sum of the
housing units (HU) impacted by surface inundations. What this data finds are the sum of the tax
lot HU’s within the water surface (Ul) and the total HU’s in each block group (Uc). The
parameters used in the dissolve tool are the census blocks for the dissolve field and the housing
units of the tax lots for the statistics field. The total population from the census data was also
added as a statistic field since it is needed later in the equation (POPc).
To determine the percentage of impacted HU’s in each block group Ul /Uc, a field was
added to each of the layers for Harris and Galveston County at each inundation level called
“Percent_HU.” The Calculate Field tool was used, using Python 3 as the expression type, and
the sum of HU’s in each water surface inundation was divided by the sum of total HU’s in each
county census block group. The next step is to solve for the POPd or the total population in each
block group. A new field is created, “POP_Derived,” and the Field Calculator tool was used to
multiply the percent of HU’s, “Percent_HU”, by the total population in each block group. This
generates the total dasymetric derived population or the POPl in each block group. The formula
for calculating this is:
POPI – POPC * UI / UC
where UI is the number of proxy units at the tax lot level (RU or ARA), UC is the number of
proxy units at the block level (RU or ARA), and POPI is the census population. (Maantay,
Maroko, and Herrmann 2007)
36
The CEDS method disaggregated the data at the tax lot level and then re-aggregated it
back to the block group level. These steps are done with five iterations for each county, one
modeling current sea-level and one for each of the four projected water surface inundation
extents for each county. The final product of the CEDS method are ten layers with a derived
impacted population within water surface inundation extents based on census and parcel level
data.
3.3.3 Modeling Vulnerability
A SoVI was created to define social vulnerabilities within each water surface elevation
for Harris and Galveston, County Texas. The deciding indicators selected are shown in Table 2
for a total of sixteen. They encompass each of the four categories of vulnerability that are
significantly associated with the population most susceptible to storm surge and flooding. The
appropriate census data was joined with each block group and was then intersected with water
surface elevations of zero-, two-, three-, four, and five-feet.
37
Table 5. Vulnerability factors and indicator selection
Vulnerability Domains Vulnerability Factors Description
Socioeconomic status Below poverty level/low
income
The past 12 months below
poverty
Unemployed Total Unemployed
No high school education Total education up to 12
th

grade with no diploma
Household composition and
disability
Elderly 65 and over
Young 5 and under
Disabled Disabled Veterans and
non-Veterans
Single parents No spouse present with
children under 18
Social identity and language Do not speak English well Combined do not speak
English well/not at all
Female Total Female
Black/African American Total Black/African
American
Asian Total Asian
Hispanic Total Hispanic
Housing and transportation Persons in group quarters Total in group quarters
Renting Total renters
No vehicle Total no vehicle
Proximity to Public
Transportation
Bus Stops

Based on Saaty’s (2008) weighting method, each vulnerability indicator was reclassified
to create a scale factor between 1-5 using census data for each specific vulnerability factor. This
was done using the field calculator and each scale factor reclassification is shown in Appendix B.
A scale factor of 5 signifies it is more favorable and of higher importance. To create a scale
factor for the proximity to public transportation, or bus stops, a buffer was created around each
block group of a quarter mile. The bus stops were scaled 1-5 as well; however, a higher weight
was given to areas with minimal or no bus stops. Each vulnerability indicator table was
38
intersected with each water surface elevation generating five new layers. Next, these five layers
were converted to raster layers using the Polygon to Raster tool with the scale factor as the value.
This creates new raster layers for each of the water surfaces and each of the sixteen
vulnerabilities with values ranging from 1-5 for a total of 80 layers.
An AHP was run to calculate the percentage or weights of each indicator. The AHP uses
a pairwise comparison of the sixteen variables to compare to each other and ranks them on a
scale of 1-9 as shown in Figure 9. A rank of 1 means the two variables are equal and 2 through 9
indicates how much weight the two variables should hold. The output of the AHP assigns each
weight a percentile rank to create an index ranking indicator set. The final index score was then
used in a weighted overlay.

Figure 11. AHP indicator ranking scale example

The weighted overlay tool was used to overlay the raster layers by measuring the weights
of each according to their importance. The sixteen rasters created in the previous step for the
water surfaces were added as the input rasters. The output percentile rank derived from the AHP
was used for the percent of influence of each indicator. The field value range of the indicators
were 1-5 based on the original scale factor. This process was repeated for each water surface
elevation. The final product results in five weighted overlay rasters indicating the areas
containing the most susceptible population.
39
3.3.4 Population Growth and Exponential Smoothing Algorithm (ETS)
The ETS algorithm is computed using Microsoft Excel with the CEDS data for 2020 and
the Brown University data from 1960 through 2010 to calculate the future population. In the
previous step, the CEDS data was intersected with the water surface elevations to show the
population within inundations. To prepare the Brown University data, each year was also
intersected with the water surface elevations to account for the same geographic population as
the CEDS method. The ETS is then performed with this historical data to project the future
population for 2050 and 2100.
The ETS computes a forecast using three required variables, Target Date, Values, and
Timeline, and three optional variables, Seasonality, Data Completion and Aggregation. The
target date is the value to be projected. For the purpose of this study, two ETS forecasts are run
to project the population using target dates of 2050 and 2100. The Values are the numeric data
that is being forecasted or the historical population from each year from 1960-2020. The
Timeline is the step between each data set. For this project the timeline is ten years because the
census data is decennial. The ETS forecast optional parameters are for Seasonality, Data
Completion and Aggregation. The Seasonality is a number that informs the algorithm whether it
should use seasonality, anything above a value of one, or if it is linear, a value of 0. The pattern
of seasonality should follow the Timeline; however, by using a value of one the formula will
auto detect the Seasonality variable (Microsoft 2021). A value of one is used to allow auto
detection as the data is straightforward and has an interval of exactly ten years. The Data
Completion value is used when Values are missing, in this case it is referring to the data from
1960-2020. A value of one interpolates that data and fills in the missing values and a value of
zero replaces the value with zero. Since there are no missing values in this dataset, this option is
not used. The last variable is Aggregation. This variable is numeric and is used if there is
40
duplicate data for the same Timeline. As an example, if the census data for 2020 and the CEDS
method data for 2020 are both used then Aggregation needs to be established. The options for
this variable are AVERAGE, SUM, COUNT, COUNTA, MIN, MAX, and MEDIAN. The
default value uses AVERAGE. For this study the CEDS data is used, as this is the more accurate
representation of the population, and the default value of one or AVERAGE. The final output
shows the projected values, or population, for the years 2050 and 2100. These values depict an
estimate of who will be impacted by water surface inundations for projected SLR in 2050 and
2100 compared to the current population established for 2020.

41
Chapter 4 Results
This study accomplished its goal of identifying the vulnerable population within SLOSH
inundations for a SLR of zero-, two-, three-, four, and five-feet to show current conditions and
SLR elevations for 2050 and 2100. The results consist of four subsections; water surface
elevations, CEDS method, vulnerability index, and projected future population. Each section
considers both Harris and Galveston County results.
4.1 Water Surface Elevations
The purpose of determining storm surge at different SLR elevations is to discover who is
within those inundations, their socioeconomic status, and vulnerable population. One important
vulnerability is being within storm surge inundation boundaries, and this is the first step in
understanding the communities in Harris and Galveston County. This was accomplished by
combining SLOSH with differing SLR elevations to create a water surface elevation. The final
water surfaces created from the merged SLR elevations and SLOSH show where MOM storm
surge with a CAT 5 hurricane at high tide will extend. The different water surfaces indicate
storm surge at current sea-level elevation, SLR zero feet, and what is likely to occur in the years
2050, SLR two- and three-feet., and 2100, SLR four- and five-feet. Figure 13 portrays a water
surface at SLR of zero feet with residential lots for Harris and Galveston County to illustrate how
storm surge affects the current population. The water surface almost entirely encompasses
Galveston Island and a part of Galveston County. If a CAT 5 storm at high tide were to strike
this area Galveston Island would be almost completely inundated. Harris County fared better
with most of the residential areas to the north and northwest. However, the areas near Galveston
42
Bay and the Houston Ship Channel are already within the water surface inundations at current
conditions.

Figure 12. Water Surface Inundations at SLR zero feet for 2020, Harris and Galveston Counties

As sea-level continues to rise, a greater population will fall within water surface
inundations. The following figures illustrate the progression of inundation as sea level rises and
the additional residential parcels affected. Each county is shown separately. Since Harris County
is very large and inundations are only near Galveston Bay and the Houston Ship Channel, Figure
1516 only shows the section of the county that is within inundations. Figure 1617 depicts
Galveston County from current conditions to SLR of five feet.
43

Figure 13. Harris County WSE at SLR zero-, two-, three-, four, and five-feet
44

Figure 14. Galveston County WSE at SLR zero-, two-, three-, four, and five-feet

As sea level rises in Harris County, the most notable rise in water surface elevations is east
of Galveston Bay and the northeast near the Houston Ship Canal, San Jacinto River, and Buffalo
Bayou. Galveston county was mostly inundated at SLR of zero, but as inundations rise to five
feet, the southeastern section near Avenue R ½ becomes submerged.
45
Defining the water surface elevations and intersecting the residential parcels is the first step
in discovering the at-risk population. The next step is establishing a SoVI to determine the
socioeconomic status of the inhabitants within these inundations.
4.2 Cadastral-based Expert Dasymetric System
Through the CEDS method of data disaggregation and reaggregation, this analysis shows
the population density for Harris and Galveston County within each water surface elevation. The
sum of the at-risk population was calculated at the tax lot level, reaggregated back to the block
level, and intersected with each water surface layer. The water surface of zero feet represents the
current conditions of the population impacted in 2020. As sea level rises the projected impacts
are shown using two- and three feet for 2050 and four- and five feet for 2100. This estimated
impact on the population is categorized by density per square mile. The density is shown by
block group within each water surface elevation.
4.2.1 CEDS Method Galveston County
Through generating a population density map for Galveston County, the areas of
vulnerability are analyzed at each SLR projection. Figure 16 shows the differences in population
density for each water surface elevation with the population density ranges. It’s noticeable that as
sea level rises and the water surface encroaches further inland, the population density increases.
46

Figure 15. Galveston County water surface population density map

Most of Galveston County is affected by inundations regardless of sea-level rise with a
few exceptions. The northeast area, the central-east area, and the northwest, become further
impacted as sea level rises to five feet. The northeast area lies on the Galveston Bay and touches
Moses Bay, Dollar Bay, and Clear Lake. This area, along with other blocks that adjoin water
bodies and rivers, are primarily affected with heightened sea-levels. The northeast area with the
47
highest densification falls within Texas City with a few blocks indicating more than 2,000 and/or
3,000 persons per square mile. Another high densification area is the central part of Galveston
County, near Santa Fe with over 2,000 persons per square mile. Galveston Island is similar
where most of the island is inundated at current conditions; however, there is a small section that
is not impacted until sea level rises to four feet as shown in Figure 18.

Figure 16. Galveston Island water surface population density maps at zero- and four-feet

48
Although the southern part of the island sits directly on the Gulf of Mexico, it is protected
by a 5-foot-wide, 17-foot-high seawall. This alleviates some inundation on the southeastern end
of the island. Although at a water surface of four feet, this area is still flooded. The northern area
of Galveston Island consists mostly of shipyards for oil and mining and cruise line docks. Since
only residential lots were considered, this area is not included in this study.
4.2.2 CEDS Method Harris County
The population density map for Harris County shows slightly different results from
Galveston County. The population density is shown in Figure 19 with the density ranges for each
water surface elevation. For Harris County the population density appears to become less once
the water surface reaches five feet. However, this does not necessarily indicate a decline in the
population within those block groups, but a larger area in square miles. For instance, the
northeastern section near Mont Belvieu (the large yellow area to the northeast) is only inundated
at five feet and covers an area of over 40 sq. mi., but only a few residential parcels are within the
water surface inundations; therefore, decreasing the population density.

49

Figure 17. Harris County water surface population density

Most of the areas inundated are adjacent to the Houston shipping Canal, Buffalo Bayou,
which extends west into downtown Houston, and the San Jacinto River, which runs to the north.
There are four main areas of concern with a population density of over 1,500. The southwest
area, west of I-45 near Friendswood north of Clear Creek River, the southeast area near
Seabrook and El Lago north of Clear Lake, the northeast area near Lynchburg, and the north
50
central area near Cloverleaf. Both Lynchburg and Cloverleaf are surrounded by numerous
streams and canals.
4.3 Vulnerability Index
The SoVI and AHP calculated the at-risk population by weighing the vulnerability
indicators within the block groups intersected with the water surface elevations. These values
were interpreted using a weighted overlay to indicate the areas of most susceptibility to hurricane
storm surge and flooding. The indicators used in the SoVI are listed in Table 2. An AHP was
calculated using an online calculator supplied by Business Performance Management Singapore
that can be found at the following URL, https://bpmsg.com/ahp/ahp-calc.php. The AHP was
given sixteen indicators for a total of 120 pairwise comparisons.
The Eigenvalue of the AHP, which indicates variance between the selected factors and
should be above 1, was 18.18%. The consistency ratio of the pairwise comparison was 9.1%
which should be below 0.1 (10%) to indicate an acceptable measure of reliability (Saatay 1990).
The final output of the AHP was within satisfactory tolerances for both eigenvalue and the
consistency ratio. The resulting AHP matrix with the final indicators is shown in Figure 21.

51

Figure 18. Analytic Hierarchical Process matrix results

The resulting percentages from the AHP were applied to a weighted overlay for each
water surface elevation. Figure 22 shows the areas with the highest vulnerability for each along
with a legend. Areas in red designate the highest vulnerable populations or a scale factor of 5
followed by the areas in orange with a scale factor of 4.
52

Figure 19. Weighted overlay vulnerability indicator results

The highest vulnerable area within the water surface elevations is northwestern Texas
City. Located in Galveston County, Texas City resides along the coast of Galveston Bay. The
northeastern section of the city, shown in orange, indicates the next highest area of vulnerability.
Other areas of interest in Galveston County are to the south near the causeway entrance to
Galveston Island, near Broadway Street and Harborside Drive, to the west of Texas City near
Hitchcock and Santa Fe, and to the northwest near League City which extends into Harris
53
County near Webster. There are no extremely vulnerable areas of interest in Harris County;
however, to the north there are two noticeable orange areas. One area is near Lynchburg, on the
coast of Burnet Bay and Buffalo Bayou near the Lynchburg Reservoir. Further north, near
Magnolia Gardens is another area of concern. This area lies east of the San Jacinto River, south
of Lake Houston, and has numerous surrounding lakes.
4.4 Population Growth and Exponential Smoothing Algorithm (ETS)
The population ETS was derived for each block group within each water surface
elevation for the projected population in 2050 and 2100 using Microsoft Excel Forecast.ETS
calculation. The Brown University data along with the CEDS method data from 1960 through
2020 was used as the historical or past data of which the forecasted values were generated. The
data shows a steady rise in population until 2010 and then a slight decline in 2020 in both
counties with a few exceptions. Table 3 reflects the population from 1960 through 2100 for
Harris County and Table 4 shows Galveston County. The data covers the population within the
water surface elevations only and not the entire counties. Harris County shows an incline in
population in 2050 and 2100 for all water surface elevations leaving even more people
vulnerable. Galveston County shows a steady incline as well, except in block three in each water
surface elevation. Block three had a noticeable decline in population starting in 2020 and the
ETS shows a continual decline into 2050 and 2100.

54
Table 6. Harris County Exponential Smoothing Algorithm, existing and forecast results
Harris 1960 1970 1980 1990 2000 2010 2020 2050 2100
WSE 0 -
Block 1 56,749 59,864 63,548 64,267 69,698 85,061 74,679 92,165
110,57
6
WSE 0 -
Block 2 57,680 59,177 60,571 61,028 62,975 66,868 68,391 73,745 83,086
WSE 0 -
Block 3 27,076 28,001 28,469 30,145 31,290 31,789 33,958 36,816 42,610
WSE 0 -
Block 4 2,050 2,235 2,415 2,648 2,719 3,218 3,008 3,900 4,785
WSE 2 -
Block 1 58,737 62,378 64,257 65,120 67,896 86,215 80,959 96,547
117,86
4
WSE 2 -
Block 2 60,019 61,494 62,467 64,927 66,006 68,164 72,565 76,953 87,218
WSE 2 -
Block 3 28,062 28,846 29,423 30,648 31,141 32,469 33,958 36,504 41,447
WSE 2 -
Block 4 2,147 2,394 2,761 3,085 3,373 4,621 5,525 7,110 10,116
WSE 3 -
Block 1 65,488 68,722 72,852 74,649 77,369 86,574 83,221 95,519
111,02
5
WSE 3 -
Block 2 61,095 62,890 64,561 66,318 68,642 71,346 74,170 82,593 96,633
WSE 3 -
Block 3 28,770 29,569 30,154 31,284 32,231 33,149 34,900 37,673 42,850
WSE 3 -
Block 4 2,289 2,427 2,986 3,214 3,618 4,826 5,684 7,321 10,373
WSE 4 -
Block 1 68,340 72,656 76,485 78,033 84,541 92,614 90,716 105,326
125,10
0
WSE 4 -
Block 2 61,797 64,865 66,567 69,287 75,307 76,168 80,155 89,550
105,27
6
WSE 4 -
Block 3 29,357 30,369 31,025 31,836 35,602 36,259 37,911 42,312 49,957
WSE 4 -
Block 4 2,401 2,648 3,287 3,537 3,948 5,148 5,863 7,562 10,619
WSE 5-
Block 1 72,638 77,053 81,071 85,253 94,127 98,165 102,823 118,705
144,76
7
WSE 5-
Block 2 63,281 66,693 68,792 71,952 75,287 77,952 80,155 88,989
103,01
8
WSE 5-
Block 3 30,186 31,153 31,869 32,404 35,604 36,846 37,911 42,018 48,943
WSE 5-
Block 4 2,563 2,862 3,573 3,822 4,325 6,084 7,214 9,433 13,575

55
Table 7. Galveston County Exponential Smoothing Algorithm, existing and forecast results
Galveston 1960 1970 1980 1990 2000 2010 2020 2050 2100
WSE 0 -
Block 1
63,557 66,212 68,176 68,860 71,724 78,497 73,031 81,753 89,957
WSE 0 -
Block 2
40,514 44,989 48,559 54,005 56,361 59,566 76,557 86,236 115,54
6
WSE 0 -
Block 3
33,961 35,737 36,543 38,113 43,105 42,666 27,474 35,745 31,494
WSE 0 -
Block 4
15,278 15,981 16,373 17,187 17,281 17,479 17,074 18,069 19,350
WSE 0 -
Block 5
2,291 2,479 3,228 3,795 4,213 4,479 3,723 5,284 6,716
WSE 0 -
Block 6
2,267 2,329 2,484 2,603 2,780 2,805 2,494 2,921 3,201
WSE 2 -
Block 1
72,082 74,409 75,736 76,509 79,430 85,274 78,848 86,805 92,737
WSE 2 -
Block 2
53,465 54,173 54,862 56,854 58,315 61,236 84,103 85,063 108,44
1
WSE 2 -
Block 3
36,812 37,679 38,527 39,084 41,151 44,819 29,416 35,223 29,975
WSE 2 -
Block 4
16,532 17,235 17,247 17,432 17,526 18,380 17,642 18,645 19,502
WSE 2 -
Block 5
2,378 2,516 3,358 3,624 4,367 4,602 3,912 5,529 6,928
WSE 2 -
Block 6
2,403 2,541 2,674 2,814 2,964 3,215 2,648 3,216 3,456
WSE 3 -
Block 1
71,889 74,202 77,364 78,267 80,216 87,301 80,338 89,418 97,009
WSE 3 -
Block 2
53,979 57,642 58,964 59,315 61,745 78,686 90,128 105,85
5
138,04
7
WSE 3 -
Block 3
37,264 38,862 39,527 40,126 43,630 47,497 32,253 39,410 36,215
WSE 3 -
Block 4
17,023 17,278 17,385 17,526 17,824 18,620 17,962 18,932 19,834
WSE 3 -
Block 5
2,842 3,167 3,408 4,136 4,430 4,724 4,023 5,319 6,392
WSE 3 -
Block6
2,407 2,566 2,713 2,898 3,047 3,346 2,941 3,534 4,014
WSE 4 -
Block 1
78,228 79,736 82,724 83,732 80,701 92,317 84,203 96,732 103,43
0
WSE 4 -
Block 2
55,167 59,128 61,342 62,020 62,605 86,207 93,114 110,99
8
145,28
7
WSE 4 -
Block 3
38,120 39,415 40,325 41,123 46,879 49,844 33,372 42,450 42,424
WSE 4 -
Block 4
17,447 17,530 18,115 17,715 16,744 19,276 17,980 18,642 19,206
56
WSE 4 -
Block 5
3,343 3,502 3,521 4,218 3,481 4,847 3,723 5,382 5,480
WSE 4 -
Block 6
2,486 2,592 2,618 3,062 3,210 3,653 2,494 3,371 3,520
WSE 5-
Block 1
81,420 83,347 87,367 88,484 86,513 96,758 90,100 102,98
1
111,25
7
WSE 5-
Block 2
59,147 62,707 63,451 63,678 64,129 95,635 94,773 110,37
6
143,74
8
WSE 5-
Block 3
39,284 41,521 42,064 44,612 49,196 52,255 34,963 44,446 41,929
WSE 5-
Block 4
17,862 18,124 18,635 17,886 18,240 19,852 17,980 18,968 19,516
WSE 5-
Block 5
3,486 3,521 3,599 4,383 3,481 4,969 3,723 5,350 5,902
WSE 5 -
Block 6
2,547 2,645 2,815 3,210 3,376 3,925 2,494 3,507 3,609

57
Chapter 5 Discussion and Conclusions
This study assessed the effects of SLR and hurricane storm surge on the community and the
vulnerable population within Harris and Galveston County, Texas. The current areas of concern
for 2020 were established as well as the projected population in 2050 and 2100 as sea level rises.
The goal was to ascertain where storm surge would encroach with rising sea levels, the
socioeconomic vulnerable peoples within that area, and the projected population within
estimated sea level rise elevations.
This chapter reviews the results of the final assessment of the inundation areas and the
population within. The study findings are discussed along with the limitations and
considerations. The final section compares the results of this study to similar study findings for a
greater understanding of the issues that Harris and Galveston County face.
5.1 Study Findings
This analysis discovered the areas of significance across Harris and Galveston County
with rising-sea- levels and storm surge. Throughout the study, there was a common theme in
some locations. Combining the CEDS population density maps, the weighted overlay, and the
water surface elevation at five feet of SLR, these patterns become apparent. Most notably, Texas
city had the highest population density, with more than 2,000 – 3,000 people per square mile,
and the highest vulnerable population in both Harris and Galveston County as indicated in Figure
24. This suggests that this is a major area of concern for evacuations and emergency
management personnel during hurricanes. Other areas that have recurring themes of high
population density and high vulnerability are the Santa Fe/Hitchcock area in Galveston County
and the Lynchburg/Channelview area in Harris County.
58

Figure 20. Density map and weighted overlay with a water surface elevation of five feet

Highly vulnerable populations that may also need assistance during hurricane flooding
are located near Broadway Street and Harborside Drive and League City/Webster in Galveston
County, and Magnolia Gardens, Cloverleaf, and Seabrook/El Lago in Harris County. The areas
with high socioeconomic vulnerability and the least available public transportation (bus stops)
are the Santa Fe/Hitchcock area with the nearest bus stop just under a mile and the League
City/Webster area with the nearest bus stop over two miles away.
The past population data indicates an overall increase in population throughout most of
the decades. The one anomaly was in 2020 when the population declined in most areas.
59
Fluctuations may be caused by population relocation and/or a decreased desire to live in coastal
communities prone to flooding and storm surge. In 2017 Hurricane Harvey stagnated over the
study area and dumped record amounts of rainfall over Houston and Galveston. The devastation
that occurred may be the source of the decline in population.
The resulting ETS shows a population growth pattern into 2050 and 2100. The one
exception to this is block three in Galveston County as mentioned in the results section. The
average percent increase in population in Harris County is 119% with block one having the
highest population rise. This block encompasses Magnolia Gardens, Cloverleaf, and part of
Seabrook where there is high socioeconomic vulnerability. This indicates that as sea level rises
and the population increases in these areas, even more people will be at risk. The average percent
increase in population in Galveston County is 115% with block two having the highest
population rise. This block includes the area near Broadway Street and Harborside Drive, a small
part of Hitchcock, League City, and an even smaller section of Texas City. This suggests that the
areas with the most growth are not a large portion of the vulnerable areas within this county.
5.2 Limitations and Considerations
This study used census data from ACS and Brown University, accredited to NHGIS. The
Brown University data ultimately came from the census bureau and ACS. There are indications
that this data is incomplete and does not accurately assess the current population. As Miyake et
al. (2010) stated it does not always incorporate the poor, homeless, undocumented immigrants,
and other marginalized peoples. The census data may exclude citizens and the final output only
represents the findings of the data available. Improvements in the census data or the collection
process would generate a more accurate representation of the population. As mentioned in the
previous section, the census data for 2020 showed a decline in most areas within both counties.
60
This may also be caused from inaccurate census data. Since the ETS was developed utilizing
census data, this inherently can cause inaccuracies within the estimated future population.
Socioeconomic vulnerable peoples within SLR and SLOSH MOM High Tide Cat 5
elevations were analyzed but did not incorporate the entire county. The data in this study only
included residents within the water surface elevations. Although this encompassed almost all of
Galveston County, a great deal of residential lots in Harris County to the north and northwest
were not within the inundations. This data also does not account for situations like Hurricane
Harvey where rainfall is a major concern and should be taken into consideration. The actual
susceptible community may signify a different population in the event of a Cat 5 hurricane with
extreme rainfall amounts; however, the data presented will give emergency management
professionals and first responders an indication of the communities in need.
The study indexing, or the indicators themselves, are based on judgement. It is founded
on the professional community’s assessment of socioeconomic vulnerability. These indicators
may exclude populations that have other incapacities and need assistance. Other geophysical
attributes, such as proximity to evacuation routes, shelters, police/fire stations, may also paint a
different picture. This technique can be used in additional studies incorporating these factors to
discover different susceptibilities.
The python script was only used for the creation of the water surface elevations but can
be extended to include the CEDS method. The script made the inundation process easier to run
for repetitive analyses. It looped through each SLR elevation to create a new raster by combining
NOAA's Sea Leve Rise rasters with NOAA's SLOSH MOM Cat 5 High Tide interpolated raster
surface. It then subtracted the DEM and set null to remove cells that are not water surface
61
elevations. This script can be expanded upon or modified and applied to any scenario that
requires looping through data.
5.3 Comparison Analysis and Conclusion
The Houston-Galveston area has been studied throughout the years due to its low-lying
coastal location, its substantial population, economics, infrastructure, and its persistent flooding.
This study focused on how SLR and storm surge affects the communities in Harris and
Galveston County and which areas have the most vulnerable population. Previous studies vary in
location with some focusing on only Houston, others Galveston, and others incorporating the
Greater Houston Area. Some indicators used in these other analyses consisted of land
subsidence, air pollution, Superfund sites, home health care centers, FEMA’s National Flood
Hazard Layer, and how they are relative to either flooding or SLR and the vulnerable
populations.
In a study from Chakraborty, Collins, and Grineski (2019) on the environmental justice
implications from Hurricane Harvey flooding, portions of their choropleth map that overlap this
projects study area shows the most flooding occurred in Lynchburg and Magnolia Park in Harris
County and Hitchcock, near Broadway Street and Harborside Drive, and League City in
Galveston County. There was also a high correlation between Harvey flooding and Black,
Hispanic, and socioeconomically deprived residents within this area. This is not inclusive of all
the areas reflected in their study, but the areas that show overlap in each of the studies.
Another study by Fucile-Sanchez and Davlasheridze (2020) discovered the socially vulnerable
population after Hurricane Ike, 2008, in Galveston County. Using similar vulnerability indicators
as this study, they found that the areas with the highest susceptible peoples were Hitchcock,
Texas City, and San Leon. A study in Houston by Bodenreider et al. (2019) on the social,
62
economic, and geographic vulnerability pre- and post-Hurricane Harvey used demographic data
and environmental factors, such as, Superfund sites, wastewater discharge, and ozone. This study
only incorporated the metropolitan areas of Houston; however, Magnolia Park and Cloverleaf are
within multiple areas of concern. Both are represented in the percent of people living in poverty
and percent of people of color in relation to Superfund sites and the percent of people living in
poverty relative to air pollution. In a comparison of the results of this study to other studies an
obvious repetitiveness is found. The most common areas are listed in Table 5.
Table 8. Repetitive Vulnerable Areas in Harris and Galveston County
Harris County Galveston County
Magnolia Park Entrance to Galveston Island (near Broadway Street
and Harborside Drive)
Cloverleaf/ Lynchburg League City/Webster
Seabrook/El Lago Hitchcock/Santa Fe

For emergency management and disaster relief purposes the above-mentioned areas
should be the main focus. The highest population increase has also been seen within these
locations and they also lack public transportation. Figure 25 shows the boundaries of these cities
from the Houston-Galveston Area Council's Regional Data Hub (H-GAC) for a geographical
perspective, except for Magnolia Park and Broadway Street and Harborside Drive into Galveston
County. The boundary for these areas is sections of the weighted overlay and population density
vulnerability areas. Magnolia Park is a section in Houston and the town/city boundary includes
the entire metropolitan area.
63

Figure 21. Geographically Significant Vulnerable Areas

This study identified the vulnerable population and provided insight into the current areas
of interest and the estimated growth patterns into 2050 and 2100. Although this study only
incorporated the communities within SLR and SLOSH Cat 5 MOM inundations or water surface
elevations, it is indicative of the areas in need from multiple other studies. The SLOSH models
64
accounted for tide levels, forward speed, storm categories, and atmospheric pressure; however,
this study can be used as a foundation and expanded upon to account for precipitation, stream
flow, subsidence, or past hurricane paths. This research gives government officials, policy
makers, and emergency managers awareness of their local communities and the population at
risk and allows Harris and Galveston County to better prepare for hurricanes and natural
disasters.
65
References
Bass, B. and Bedient P. 2018. “Surrogate Modeling of Joint Flood Risk Across Coastal
Watersheds.” Journal of Hydrology 558: 159-173.
Baykal, T. M., Colak, H. E., and Kılınc, C. “Forecasting Future Climate Boundary Maps (2021–
2060) Using Exponential Smoothing Method and GIS.” The Science of the total
environment 848 (2022): 157633–157633.
Blake, E. S. 2018. "The 2017 Atlantic Hurricane Season: Catastrophic Losses and Costs."
Weatherwise 71, no. 3: 28-37.
Blessing, R., Sebastian, A., and Brody, S. D. 2017. “Flood Risk Delineation in the United States:
How Much Loss Are We Capturing?” Natural Hazards Review 18, Issue 3.
Bodenreider, C., Wright, L., Barr, O., Xu, K. and Wilson, S.. 2019. “Assessment of Social,
Economic, and Geographic Vulnerability Pre- and Post-Hurricane Harvey in Houston,
Texas.” Environmental justice 12, no. 4 (2019): 182–193. American Society of Civl
Engineers. DOI: 10.1061/ (ASCE)NH.1527-6996.0000242
Boyer, D., and Vardy, M. “Flooded City: Affects of (Slow) Catastrophe in Post-Harvey
Houston.” Current anthropology 63, no. 6 (2022): 615–636.
Burton, C. G. 2010. "Social Vulnerability and Hurricane Impact Modeling." Natural hazards
review 11, no. 2: 58-68.
Carlson, C., Goldman, G., Dahl, K. 2016. Stormy Seas, Rising Risks: Assessing Undisclosed
Risk from Sea Level Rise and Storm Surge at Coastal US Oil Refineries. In: Drake, J.,
Kontar, Y., Eichelberger, J., Rupp, T., Taylor, K. (eds) Communicating Climate-Change
and Natural Hazard Risk and Cultivating Resilience. Advances in Natural and
Technological Hazards Research, vol 45. Springer, Cham. https://doi.org/10.1007/978-3-
319-20161-0_19
Chakraborty, J., Collins, T. W., and Grineski, S. E. 2019. "Exploring the Environmental Justice
Implications of Hurricane Harvey Flooding in Greater Houston, Texas." American
journal of public health 109, no. 2: 244-250.
Conrad, H. “Are Coastal Marshes Drowning Faster than Expected?” Texas A&M University
Engineering, June 17, 2021. https://engineering.tamu.edu/news/2021/06/ocen-are-coastal-
marshes-drowning-faster-than-expected.html.
Cutter, S. L., Boruff, B. J., and Shirley, W. L. 2003. "Social Vulnerability to Environmental
Hazards." Social science quarterly 84, no. 2: 242-261.
Cutter, S. L., Mitchell, J. T., and Scott, M. S. 2000. "Revealing the Vulnerability of People and
Places: A Case Study of Georgetown County, South Carolina." Annals of the association
of American Geographers 90, no. 4: 713-737. “Coastal Areas.” The United States Census
66
Bureau, July 23, 2020. https://www.census.gov/topics/preparedness/about/coastal-
areas.html.
Davis Jr, A. B. 1951. "History of the Galveston Sea Wall." Coastal Engineering Proceedings, no.
2: 24-24.
Federal Emergency Management Administration (FEMA) (2018). FEMA - Harvey Flood Depths
Grid, HydroShare, https://doi.org/10.4211/hs.165e2c3e335d40949dbf501c97827837
Flanagan, B. E., et al. 2011. "A Social Vulnerability Index for Disaster Management." Journal of
homeland security and emergency management 8, no. 1.
Fucile-Sanchez, E., and Davlasheridze, M. 2020. “Adjustments of Socially Vulnerable
Populations in Galveston County, Texas USA Following Hurricane Ike.” Sustainability
(Basel, Switzerland) 12, no. 17 (2020): 7097–.
Galloway, D. L., Jones, D. R., and Ingebritsen, S. E. 1999. Land Subsidence in the United States.
Vol. 1182: US Geological Survey.
Gardner, E. S. 2006 “Exponential Smoothing: The State of the art—Part II.” International journal
of forecasting 22, no. 4 (2006): 637–666.
Gathongo, N., and Tran L.. “Assessing Social Vulnerability of Villages in Mt. Kasigau Area,
Kenya, Using the Analytical Hierarchy Process.” GeoJournal 85, no. 4 (2020): 995–1007.
“Geo Galveston - Texas.” Temperature - Precipitation - Sunshine - Snowfall, 2021. US Climate
Data. Accessed September 20, 2021. https://www.usclimatedata.com/map/USTX0499.
Harris County Flood Control. 2021. Hurricane Ike 2008. Accessed September 20, 2021.
https://www.hcfcd.org/About/Harris-Countys-Flooding-History/Hurricane-Ike-2008
Horton, B. P., et al. 2020. "Estimating Global Mean Sea-Level Rise and Its Uncertainties by
2100 and 2300 from an Expert Survey." npj Climate and Atmospheric Science 3, no. 1:
1-8.
Houston Harris County Winter Storm Relief Fund. 2023. “Disproportionately Impacted
Communities.” Accessed January 14, 2023.
https://winterstormrelieffund.org/?page_id=131
“Houston Population 2022.” Population USA. Accessed March 10, 2022.
http://www.usapopulation.org/houston-population/
IPUMS USA, University of Minnesota, www.ipums.org. Accessed November 13, 2010.
Kasperson, J. X., Kasperson, R. E., and Turner, B. L. 1996. "Regions at Risk: Exploring
Environmental Criticality." Environment: Science and Policy for Sustainable
Development 38, no. 10: 4-29.
67
Logan, J. R, and Xu Z. 2015. “Vulnerability to Hurricane Damage on The US Gulf Coast Since
1950.” Geographical review vol. 105,2 (2015): 133-155. doi:10.1111/j.1931-
0846.2014.12064.x
Maantay, J. A., Maroko, A. R., and Herrmann, C. 2007. "Mapping Population Distribution in the
Urban Environment: The Cadastral-Based Expert Dasymetric System (Ceds)."
Cartography and Geographic Information Science 34, no. 2: 77-102.
Maloney, M. C., and Preston, B.L. 2014. “A Geospatial Dataset for US Hurricane Storm Surge
and Sea-Level Rise Vulnerability: Development and Case Study Applications.” Climate
risk management 2, no. C (2014): 26–41.
Microsoft. 2021. FORESCASTS.ETS Function. Accessed January 7, 2023.
https://support.microsoft.com/en-us/office/forecast-ets-function-15389b8b-677e-4fbd-
bd95-21d464333f41
Miyake, K. K., et al. 2010. "Not Just a Walk in the Park: Methodological Improvements for
Determining Environmental Justice Implications of Park Access in New York City for
the Promotion of Physical Activity." Cities and the Environment 3, no. 1: 1.
NASA. April 2021. Global Climate Change. Vital Signs of the Planet. Accessed September 26,
2021. https://climate.nasa.gov/vital-signs/sea-level/
National Oceanic and Atmospheric Administration (NOAA) Climate.gov (2021) Climate
Change: Global Sea Level. Accessed September 20, 2021.
https://www.climate.gov/news-features/understanding-climate/climate-change-global-
sea-level
National Oceanic and Atmospheric Administration (NOAA) Coastal Services Center. (2012)
Mapping Coastal Inundation Primer. Accessed September 20, 2021.
https://coast.noaa.gov/data/digitalcoast/pdf/coastal-inundation-guidebook.pdf.
National Oceanic and Atmospheric Administration (NOAA) National Westher Service. Saffir-
Simpson Hurricane Scale. Accessed May 3, 2023.
https://www.weather.gov/mfl/saffirsimpson#:~:text=The%20Saffir%2DSimpson%20Hur
ricane%20Wind,loss%20of%20life%20and%20damage.
National Oceanic and Atmospheric Administration (NOAA) Sea, Lake, and Overland Surges
form Hurricanes (SLOSH) National Hurricane Center. Accessed September 20, 2021.
https://www.nhc.noaa.gov/surge/slosh.php.
National Oceanic and Atmospheric Administration (NOAA) Office for Coastal Management.
Sea Level Rise Data: 1-10 ft Sea Level Rise Inundation Extent. Accessed September 22,
2021. https://www.fisheries.noaa.gov/inport/item/48106
National Oceanic and Atmospheric Administration (NOAA) Tides and Currents. Tidal Datums.
Accessed May 3, 2023.
68
https://tidesandcurrents.noaa.gov/datum_options.html#:~:text=MHHW*,the%20National
%20Tidal%20Datum%20Epoch.
Office for Coastal Management, 2023: NOAA Office for Coastal Management Sea Level Rise
Data: 1-10 ft Sea Level Rise Inundation Extent from 2010-06-15 to 2010-08-15. NOAA
National Centers for Environmental Information,
https://www.fisheries.noaa.gov/inport/item/48106.
Oppenheimer, M., et. al. 2019. Sea Level Rise and Implications for Low-Lying Islands, Coasts
and Communities. IPCC Special Report on the Ocean and Cryosphere in a Changing
Climate [H.-O. Pörtner, D.C. Roberts, V. Masson-Delmotte, P. Zhai, M. Tignor, E.
Poloczanska, K. Mintenbeck, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama,
N.M. Weyer (eds.)]. In press.
Rahmstorf, S. (2012) Modeling Sea Level Rise. Nature Education Knowledge 3(10):4. Accessed
September 6, 2021. https://www.nature.com/scitable/knowledge/library/modeling-sea-
level-rise-25857988/
Saaty, T. L. (2008). Decision making with the analytic hierarchy process. International Journal of
Services Sciences, 1(1), 83–98.
Saaty, T. L. 1990. “How to Make a Decision: The Analytic Hierarchy Process.” European
journal of operational research 48, no. 1 (1990): 9–26.
Smiley, K. T. 2020. “Outdated and Inaccurate, FEMA Flood Maps Fail to Fully Capture Risk.”
Rice University, Kinder Institute for Urban Research. Accessed December 30, 2022.
https://kinder.rice.edu/urbanedge/outdated-and-inaccurate-fema-flood-maps-fail-fully-
capture-risk
SolarWinds. 2019. Holt-Winters Forecasting and Exponential Smoothing Simplified. Accessed
January 7, 2023. https://orangematter.solarwinds.com/2019/12/15/holt-winters-
forecasting-
simplified/#:~:text=What%20Is%20the%20Holt%2DWinters,cyclical%20repeating%20p
attern%20(seasonality).
Tate, E. 2012. “Social Vulnerability Indices: A Comparative Assessment Using Uncertainty and
Sensitivity Analysis.” Natural hazards (Dordrecht) 63, no. 2 (2012): 325–347.
Tate, E. 2013. “Uncertainty Analysis for a Social Vulnerability Index.” Annals of the
Association of American Geographers 103, no. 3 (2013): 526–43.
http://www.jstor.org/stable/23485404.
The Nature Conservatory. 2013. “Sea Level Rise. Research & Scenarios for a Changing Coast.”
StormSmart.org. Accessed January 12, 2023. http://slr.stormsmart.org/texas/
The Ocean Conference. Factsheet: People and Oceans. New York, June 5-9, 2017. United
Nations. Accessed September 8, 2021. https://www.un.org/sustainabledevelopment/wp-
content/uploads/2017/05/Ocean-fact-sheet-package.pdf
69
Turner, B. L., et al. 2003. "A Framework for Vulnerability Analysis in Sustainability Science."
Proceedings of the national academy of sciences 100, no. 14: 8074-8079.
Understanding Houston. 2023. “Vulnerability to and Impacts from Disasters. Exploring unequal
effects of natural disasters on Houston communities.” Accessed January 14, 2023.
https://www.understandinghouston.org/topic/disasters/vulnerability-
impacts#social_vulnerability
US Census. History. Cesnus Records. Accessed May 4, 2023.
https://www.census.gov/history/www/genealogy/decennial_census_records/census_recor
ds_2.html
US Department of Commerce and National Oceanic and Atmospheric Administration.
"Hurricanes in History." National Hurricane Center. Accessed September 22, 2021.
https://www.nhc.noaa.gov/outreach/history/
US Energy Information Administration. (EIA) 2021. “The Port District of Houston-Galveston
Became a Net Exporter of Crude Oil in April.” Today in Energy. Accessed September
15, 2021. https://www.eia.gov/todayinenergy/detail.php?id=36932.
US Geological Survey. About 3DEP Products and Services. Accessed May 4, 2023.
https://www.usgs.gov/3d-elevation-program/about-3dep-products-services
US Geological Survey, 2018. Characterization of Peak Streamflows and Flood Inundation of
Selected Areas in Southeastern Texas and Southwestern Louisiana from the August and
September 2017 Flood Resulting from Hurricane Harvey. by Watson, K. M., et al.
Visit Houston Texas. Facts and Figures. 2021. Accessed August 27, 2021. Houston Facts &
Figures | Find Population, Culture & Industry Data (visithoustontexas.com)
Winters, P. R. 1960. “Forecasting Sales by Exponentially Weighted Moving Averages.”
Management Science 6, no. 3 (1960): 324–42. http://www.jstor.org/stable/2627346.
Yuan, S., Guo, J., and Zhao, X. “Integrated Weighting Technique for Coastal Vulnerability to
Storm Surges.” Journal of Coastal Research, 2017, 6–12.
http://www.jstor.org/stable/44252799.

70
Appendices
Appendix A. Python Script
Creation of water surface elevations.

71

72

73
Appendix B. Scale Factor Reclassification
Field calculator calculations for each scale factor reclassification

Below poverty level/low income:

reclass(!Total_Poverty!)
def reclass(Total_Poverty):
if (Total_Poverty <= 49):
return 1
elif (Total_Poverty >= 50 and Total_Poverty <=99):
return 2
elif (Total_Poverty >= 100 and Total_Poverty <= 149):
return 3
elif (Total_Poverty >= 150 and Total_Poverty <= 199):
return 4
elif (Total_Poverty >= 200):
return 5

Unemployed:

reclass(!Total_Unemployed_1!)
def reclass(Total_Unemployed_1):
if (Total_Unemployed_1 <= 49):
return 1
elif (Total_Unemployed_1 >= 50 and Total_Unemployed_1 <=99):
return 2
elif (Total_Unemployed_1 >= 100 and Total_Unemployed_1 <= 149):
return 3
elif (Total_Unemployed_1 >= 150 and Total_Unemployed_1 <= 199):
return 4
elif (Total_Unemployed_1 >= 200):
return 5

No high school education:
reclass(!Total_noDiploma!)
def reclass(Total_noDiploma):
if (Total_noDiploma <= 199):
return 1
74
elif (Total_noDiploma >= 200 and Total_noDiploma <=299):
return 2
elif (Total_noDiploma >= 300 and Total_noDiploma <= 399):
return 3
elif (Total_noDiploma >= 400 and Total_noDiploma <= 599):
return 4
elif (Total_noDiploma >= 600):
return 5

Elderly (65 over):
reclass(!Total_65!)
def reclass(Total_65):
if (Total_65 <= 199):
return 1
elif (Total_65 >= 200 and Total_65 <=299):
return 2
elif (Total_65 >= 300 and Total_65 <= 399):
return 3
elif (Total_65 >= 400 and Total_65 <= 499):
return 4
elif (Total_65 >= 500):
return 5

Young (under 5):

reclass(!Total_Und_5!)
def reclass(Total_Und_5):
if (Total_Und_5 <= 99):
return 1
elif (Total_Und_5 >= 100 and Total_Und_5 <= 199):
return 2
elif (Total_Und_5 >= 200 and Total_Und_5 <= 299):
return 3
elif (Total_Und_5 >= 300 and Total_Und_5 <= 399):
return 4
elif (Total_Und_5 >= 400):
return 5

Disabled:

reclass(!Total_Disability!)
75
def reclass(Total_Disability):
if (Total_Disability <= 199):
return 1
elif (Total_Disability >= 200 and Total_Disability <=299):
return 2
elif (Total_Disability >= 300 and Total_Disability <= 399):
return 3
elif (Total_Disability >= 400 and Total_Disability <=499):
return 4
elif (Total_Disability >= 500):
return 5

Single parents:

reclass(!Total_SingleParents!)
def reclass(Total_SingleParents):
if (Total_SingleParents <= 49):
return 1
elif (Total_SingleParents >= 50 and Total_SingleParents <=99):
return 2
elif (Total_SingleParents >= 100 and Total_SingleParents <= 149):
return 3
elif (Total_SingleParents >= 150 and Total_SingleParents <=199):
return 4
elif (Total_SingleParents >= 200):
return 5

Do not speak English well:

reclass(!Total_Eng!)
def reclass(Total_Eng):
if (Total_Eng <= 199):
return 1
elif (Total_Eng >= 200 and Total_Eng <=299):
return 2
elif (Total_Eng >= 300 and Total_Eng <= 399):
return 3
elif (Total_Eng >= 400 and Total_Eng <=499):
return 4
elif (Total_Eng >= 500):
return 5

76

Female:

reclass(!Total_Female!)
def reclass(Total_Female):
if (Total_Female <= 399):
return 1
elif (Total_Female >= 400 and Total_Female <=699):
return 2
elif (Total_Female >= 700 and Total_Female <= 999):
return 3
elif (Total_Female >= 1000 and Total_Female <= 1499):
return 4
elif (Total_Female >= 1500):
return 5

Black/African American:

reclass(!Total_Black_AfricanAmerican!)
def reclass(Total_Black_AfricanAmerican):
if (Total_Black_AfricanAmerican<=199):
return 1
elif (Total_Black_AfricanAmerican>= 200 and Total_Black_AfricanAmerican<=299):
return 2
elif (Total_Black_AfricanAmerican>= 300 and Total_Black_AfricanAmerican<= 499):
return 3
elif (Total_Black_AfricanAmerican>= 500 and Total_Black_AfricanAmerican<= 699):
return 4
elif (Total_Black_AfricanAmerican>= 700):
return 5

Asian:

reclass(!Total_Asian!)
def reclass(Total_Asian):
if (Total_Asian <= 199):
return 1
elif (Total_Asian >= 200 and Total_Asian <=299):
return 2
elif (Total_Asian >= 300 and Total_Asian <= 399):
return 3
77
elif (Total_Asian >= 400 and Total_Asian <=499):
return 4
elif (Total_Asian >= 500):
return 5

Hispanic:

reclass(!Total_Hispanic!)
def reclass(Total_Hispanic):
if (Total_Hispanic <= 399):
return 1
elif (Total_Hispanic >= 400 and Total_Hispanic <=699):
return 2
elif (Total_Hispanic >= 700 and Total_Hispanic <= 999):
return 3
elif (Total_Hispanic >= 1000 and Total_Hispanic <= 1499):
return 4
elif (Total_Hispanic >= 1500):
return 5

Persons in group quarters:

reclass(!Total_GroupQuaters_1!)
def reclass(Total_GroupQuaters_1):
if (Total_GroupQuaters_1 <= 49):
return 1
elif (Total_GroupQuaters_1 >= 50 and Total_GroupQuaters_1 <=99):
return 2
elif (Total_GroupQuaters_1 >= 100 and Total_GroupQuaters_1 <= 149):
return 3
elif (Total_GroupQuaters_1 >= 150 and Total_GroupQuaters_1 <= 399):
return 4
elif (Total_GroupQuaters_1 >= 400):
return 5

Renters:

reclass(!Total_Renters!)
def reclass(Total_Renters):
if (Total_Renters <= 999):
78
return 1
elif (Total_Renters >= 1000 and Total_Renters <=1999):
return 2
elif (Total_Renters >= 2000 and Total_Renters <= 2999):
return 3
elif (Total_Renters >= 3000 and Total_Renters <=9999):
return 4
elif (Total_Renters >= 10000):
return 5

No vehicle:

reclass(!Total_NoVehicles!)
def reclass(Total_NoVehicles):
if (Total_NoVehicles <= 19):
return 1
elif (Total_NoVehicles >= 20 and Total_NoVehicles <=39):
return 2
elif (Total_NoVehicles >= 40 and Total_NoVehicles <= 79):
return 3
elif (Total_NoVehicles >= 80 and Total_NoVehicles <=99):
return 4
elif (Total_NoVehicles >= 100):
return 5

Proximity to Public Transportation (Bus Stops):

reclass(!BusStops!)
def reclass(BusStops):
if (BusStops ==0):
return 5
elif (BusStops >= 0 and BusStops <=5):
return 4
elif (BusStops >= 6 and BusStops <= 10):
return 3
elif (BusStops >= 11 and BusStops <= 20):
return 2
elif (BusStops >= 21):
return 1

Abstract (if available)

Abstract Communities in the Houston-Galveston area of Texas are consistently at risk of hurricane devastation. With warming climates and increasing greenhouse gases, sea-level rise (SLR) has become a significant consideration. Many studies have shown the correlation between SLR and vulnerability, however, little has been found on the implications of SLR with the influence of storm surge on the community. This study established the current population and projected future population at risk in 2050 and 2100 from SLR and storm surge inundation in Houston and Galveston County. The National Oceanic and Atmospheric Administration’s (NOAA) projections of SLR of two-, three-, four-, and five-feet are combined with NOAA’s Sea, Lake, and Overland Surges (SLOSH) predictions to produce water surface elevations as sea level rises. A social vulnerability index was created, and weights were determined, using an analytic hierarchical process to reveal the socioeconomic vulnerable population within each water surface elevation produced. A cadastral-based expert dasymetric system method was employed to improve upon census data alone for spatial data of the population in 2020. An exponential smoothing algorithm was then used to predict future populations utilizing census data from Brown University and the American Community Survey from 1960 through 2020. The final assessment establishes inhabitants who were at risk in 2020 and the projected population in 2050 and 2100 within rising sea-levels. The results identify the neighborhoods within Harris and Galveston County that are vulnerable to sea-level rise and storm surge inundation currently and in the future. This provides these two counties, and other government agencies, a geospatial assessment of vulnerable demographics within their locality and future estimates to assist in planning, preparation, and emergency response.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Mapping future population impacts caused by sea level rise in Huntington Beach and Newport Beach: comparing the cadastral-based dasymetric system to past dasymetric mapping methods

PDF

Coastal vulnerability assessment for archaeological sites on San Clemente Island and San Nicolas Island, California

PDF

Creating a flood vulnerability index for Houston, Texas

PDF

Validating the HAZUS coastal surge model for Superstorm Sandy

PDF

Assessing the use of normalized difference chlorophyll index to estimate chlorophyll-A concentrations using Landsat 5 TM and Landsat 8 OLI imagery in the Salton Sea, California

PDF

A critical assessment of the green sea turtle central west Pacific distinct population segment utilizing maxent modeling on nesting site locations

PDF

The impact of severe coastal flooding on economic recovery disparities: a study of New Jersey communities following Hurricane Sandy

PDF

Temporal analysis of soil degradation in San Joaquin County, California: a close examination of soil erosion using RUSLE

PDF

Modeling potential impacts of tsunamis on Hilo, Hawaii: comparison of the Joint Research Centre's SCHEMA and FEMA’s HAZUS inundation scenarios

PDF

Estimating populations at risk in data-poor environments: a geographically disaggregated analysis of Boko Haram terrorism 2009-2014

PDF

Geographic local routing for social connection: a novel application for the integration of routing technology and multi-user environments

PDF

Trojan Food Finder: a web-based GIS campus food sharing application

PDF

Exploring the pernicious effects of redlining and discriminatory policies on an American city: a spatio-temporal case study of New York City

PDF

Risk analysis and assessment of non‐ductile concrete buildings in Los Angeles County using HAZUS‐MH

PDF

A spatial analysis of beef production and its environmental and health impacts in Texas

PDF

Finding the green in greenspace: an examination of geospatial measures of greenspace for use in exposure studies

PDF

The geographic connotations of reincarceration: a spatial analysis of recidivism in Washington State

PDF

Measuring seasonal variation in food access: a case study of Everett, Washington

PDF

Urban green space accessibility and environmental justice: a GIS-based analysis in the city of Phoenix, Arizona

PDF

Creating a well database and web mapping application: using geographic information systems to manage and monitor groundwater resources in Sonoma County, California

Asset Metadata

Creator Seymour, Susan M. (author)

Core Title Projecting vulnerability: a combined analysis of sea-level rise, hurricane inundation, and social vulnerability in Houston-Galveston, Texas

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science

Degree Conferral Date 2023-08

Publication Date 06/01/2023

Defense Date 05/11/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Galveston,Houston,hurricane,hurricane inundation,inundation,OAI-PMH Harvest,Population,projecting,sea-level rise,SLR,socioeconomic vulnerability,Texas,vulnerability

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sedano, Elisabeth (committee member)

Creator Email smseymou@usc.edu,susan.m.seymour4@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113144557

Unique identifier UC113144557

Identifier etd-SeymourSus-11915.pdf (filename)

Legacy Identifier etd-SeymourSus-11915

Document Type Thesis

Format theses (aat)

Rights Seymour, Susan M.

Internet Media Type application/pdf

Type texts

Source 20230602-usctheses-batch-1051 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu