Development and Testing of a GIS-Based Method to Detect and Characterize Sinkholes

With the development of modern remote-sensing technology, high-resolution remotely sensed data can be used to provide more accurate mapping and analyses of surface terrain including sinkholes. The focus of this thesis is on developing a method for detecting and characterizing sinkholes including identifying particular sinkholes with the potential of having preserved paleoflood evidence. The proposed method is applied to the case study area within the Upper Tennessee River Basin, Knox County, Tennessee. The results of the method identified 7,248 likely sinkholes within the county. Based on the accuracy assessment, with a true positive rate of 84%, the results of the case study suggest that this method can be used to identify sinkholes in karst regions. Not only can the method be used to successfully identify sinkholes, but it also has several strengths in terms of sinkhole detection compared to traditional methods that rely on outdated topographic maps, aerial imagery, and field observations.