Development of Sinkhole Susceptibility Map of East Central Florida

Abstract

This paper presents the implementation of the frequency ratio (FR) method to evaluate the contribution of every factor on sinkholes and develop a sinkhole susceptibility map for East Central Florida (ECF). The FR is a probabilistic method used to determine the correlation between sinkhole locations in the past and each sinkhole-related factor. ECF, vulnerable to sinkhole formation, was selected, and a sinkhole database (total 1,018 sinkholes) was adopted from the Florida Subsidence Incident Reports (FLSIRs). The sinkhole susceptibility model was developed using a randomly selected 70% (713) of the sinkholes, with the remaining 30% (305) of the sinkholes used for validation. Seven key contributing factors were selected to calculate the sinkhole susceptibility index (SSI). The factors include hydraulic head difference, recharge rate, soil permeability, overburden thickness, aquitard layer thickness, depth to water table, and proximity to karst features. The proposed susceptibility map well predicts the relative risk of sinkhole occurrences; approximately 91.7% of reported sinkhole data falls into either very high or high susceptibility of the map.