Graduation Year

2014

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Geography

Major Professor

Jayajit Chakraborty, Ph.D.

Committee Member

Pratyusha Basu, Ph.D.

Committee Member

M. Martin Bosman, Ph.D.

Committee Member

Timothy W. Collins, Ph.D.

Committee Member

Barnali Dixon, Ph.D.

Committee Member

Sara E. Grineski, Ph.D.

Keywords

coastal amenities, coastal hazards, dasymetric mapping, GIS, social vulnerability, spatial interpolation

Abstract

While environmental justice (EJ) research in the U.S. has traditionally focused on inequities in the distribution of technological hazards, the disproportionate impacts of Hurricane Katrina on racial minorities and socioeconomically disadvantaged households have prompted researchers to investigate the EJ implications of natural hazards such as flooding. Recent EJ research has also emphasized the need to examine social inequities in access to environmental amenities. Unlike technological hazards such as air pollution and toxic waste sites, areas exposed to natural hazards such as hurricanes and floods have indivisible amenities associated with them. Coastal property owners are exposed to flood hazards, but also enjoy water views and unhampered access to oceans and the unique recreational opportunities that beaches offer. Conversely, dense urban development and associated impervious surfaces increase likelihood of floods in inland areas which may lack the amenities of proximity to open water.

This dissertation contributes to the emerging literature on EJ and social vulnerability to natural hazards by analyzing racial, ethnic, and socioeconomic inequities in the distribution of flood risk exposure in the Miami Metropolitan Statistical Area (MSA), Florida--one of the most hurricane-prone areas in the world and one of the most ethnically diverse MSAs in the U.S. The case study evaluates the EJ implications of residential exposure to coastal flood risk, inland flood risk, and no flood risk, in conjunction with coastal water related amenities, using geographic information science (GIS)-based techniques and logistic regression modeling to estimate flood risk exposure. Geospatial data from the Federal Emergency Management Agency (FEMA) are utilized to delineate coastal and inland 100-year flood hazard zones. Socio-demographic variables previously utilized in EJ research are obtained from tract level data published in the 2010 census and 2007-2011 American Community Survey five-year estimates. Principal components analysis is employed to condense several socio-demographic attributes into two neighborhood deprivation indices that represent economic insecurity and instability, respectively. Indivisible coastal water related amenities are represented by control variables of percent seasonal homes and proximity to public beach access sites. Results indicate that racial/ethnic minorities and those with greater social vulnerability based on the neighborhood deprivation indices are more likely to reside in inland flood zones and areas outside 100-year flood zones, while residents in coastal flood zones are disproportionately non-Hispanic White. Moreover, residents exposed to coastal flood risk tend to live in areas with ample coastal water related amenities, while racial/ethnic minorities and individuals with higher neighborhood deprivation who are exposed to inland flood risk or no flood risk reside in areas without coastal water related amenities. This dissertation elucidates the importance of EJ research on privilege and access to environmental amenities in conjunction with environmental hazards because areas exposed to natural hazards are likely to offer indivisible benefits.

Estimating people and places exposed to hazards for EJ research becomes difficult when the boundaries of census areal units containing socio-demographic data do not match the boundaries of hazard exposure areas. This challenge is addressed with an application of dasymetric spatial interpolation using GIS-based techniques to disaggregate census tracts to inhabited parcels. Several spatial interpolation methods are assessed for relative accuracy in estimating population densities for the Miami MSA, and the output units from the most accurate method are employed in EJ regression analyses. The dasymetric mapping efforts utilized herein contribute to research on the modifiable areal unit problem (MAUP) and its effects on statistical analyses. Since the dasymetric mapping technique used for EJ analyses disaggregates census tracts to the inhabited parcel level, the results of the associated analyses for flood hazards exposure and access to coastal water related amenities should be more reliable than those based on tracts. The enhanced accuracy associated with inhabited parcels is a result of using a more precise geospatial depiction of residential populations, which leads to a more accurate portrayal of disproportionate exposure to flood hazards. Consequently, this dissertation contributes methodologically to GIS-based techniques of dasymetric spatial interpolation and empirically to EJ analysis of flood hazards with indivisible coastal water related amenities.

Download

Share

COinS