ABSTRACT OF THE DISSERTATION 
Acute Chemical Disasters 
and Rural United States Hazardscapes 
Dissertation 9irector 
Professor Susan L. Cutter 


This dissertation examines the distribution and nature 
of acute airborne chemical releases in rural America through 
a case study approach. Two hypotheses provide the basic 
analytical framework forthe research. First, are risks and 
hazards evenly distributed in rural counties, or is there 
some pattern (clustered, regularly dispersed) to the 
distribution of incidents? Second, are there aspects of the 
local conditions of risk and response capability which 
influence the pattern of incidents and the resulting 
hazardscape? 


Four states were chosen for case studies (Alabama, 
Georgia, Iowa, and New York) based on a nationally derived 
typology of rural counties. Data were gathered on 262 
unplanned gaseous releases of hazardous materials which 
occurred from January 1982 to December 1989 in the 307 rural 
counties of the four states. For each incident, the 
information recorded included: time, place, and material 
released. County-level information included the amount of 
hazardous material present, demographic characteristics, and 
the level of local emergency preparedness and response. 
Univariate and multivariate statistical analyes, 
particularly discriminant analysis, were used to examine the 


relationship between incidents and the risk and preparedness 
context in which they occurred. 


Most incidents took place in Iowa or New York; resulted 
from fixed facilities such as chemical manufacturing plants: 
released acids, ammonia, or chlorine; and occurred in 
counties next to metropolitan areas. Several counties in 
each state experienced many more incidents than others. The 
incidents were randomly distributed in Alabama and Georgia, 
but clustered in Iowa and New York. 


A close association existed between the number of 
incidents in a county and the number of local facilities 
with hazardous materials. The higher the frequency of 
incidents, the greater the level of county emergency 
preparedness and response. The conditions in counties where 
the largest incidents occurred did not differ significantly 
from the conditions in counties where smaller incidents 
occurred. Transportation infrastructure was rarely an 
important variable. The presence of areas with higher- or 
lower-than-expected hazardousness, taken together with the 
rest of the counties, suggests that local patterns of risk 
and preparedness overlap to form a complex patchwork of 
landscape hazardousness. 


iii