In Karst areas, sinkholes are a major geohazard that creates damage and economic loss to building and civil infrastructure. The main goal of the proposed study is to develop a probabilistic spatio-magnitude sinkhole susceptibility model that predicts the potential of sinkhole occurrence on a regional scale. The dissertation presents statistical techniques such as frequency ratio (FR) and logistic regression (LR), and a machine learning technique such as artificial neural network (ANN) to develop the sinkhole hazard models. The study area was limited to the East Central Floria (ECF) region. Seven hydrogeological factors were identified as key contributing factors to sinkhole occurrence and they were used as input variables to the sinkhole hazard model. The seven factors include hydraulic head difference between surficial and upper Florida aquifers, groundwater recharge rate to the upper Floridan aquifer, soil permeability, overburden thickness, surficial aquifer system (SAS) thickness, intermediate aquifer system (IAS) thickness, and proximity to karst features. Both FR and LR were used to construct the spatial prediction models for sinkhole susceptibility mapping. In addition, ANN was employed to develop a probabilistic spatio-magnitude sinkhole hazard model. The prediction accuracy of all models was validated by area under the receiver operating characteristic (ROC) curve (AUC) analysis. Subsequently, the probabilities of location and size of the sinkhole are computed and mapped on a regional scale by Geographical Information System (GIS) techniques.
If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu.
Nam, Boo Hyun
Doctor of Philosophy (Ph.D.)
College of Engineering and Computer Science
Civil, Environmental, and Construction Engineering
Length of Campus-only Access
Doctoral Dissertation (Campus-only Access)
Kim, Yong Je, "Probabilistic Spatio-Magnitude Sinkhole Hazard Analysis for East Central Flrodia" (2019). Electronic Theses and Dissertations, 2004-2019. 6878.