Most methods for the assessment of sinkhole hazard susceptibility are predicated upon knowledge of pre-existing closed depressions in karst areas. In the United States (U.S.), inventories of existing karst depressions are piecemeal, and are often obtained through inconsistent methodologies applied at the state or county level and at various scales. Here, we present a first attempt at defining a karst closed depression inventory across the conterminous U.S. using a common methodology. Automated algorithms for extraction of closed depressions from 1/3 arc-second (approximately 10 m resolution) National Elevation Dataset (NED) were run on the U.S. Geological Survey (USGS) “Yeti” high-performance computing cluster. The full NED was first conditioned to reduce the creation of artificial closed depressions by breaching digital dams at road and stream crossings, using the flowlines and transportation route vectors from the USGS National Map. The resulting depressions were selected according to location within geologic units having the potential for karst, and screened for occurrence in areas of developed land, open water and wetlands, and areas of glacial and alluvial sediment cover. The results were used as the input to create a nationwide depression density map. Our results were compared with karst depression density maps for diverse karst regions within states that have existing closed depression inventories. The individual state-scale maps compared favorably to the results obtained from the method applied universally across the nation and illustrated regional sinkhole hotspots in known areas of well-developed karst. Limitations of the automated method includes false positive depressions resulting from artifacts generated during the computer processing of the elevation models, and inclusion of depressions resulting from non-karst geomorphic processes. More thorough examination of the screening criteria for depressions is required.
|Title||Progress toward a preliminary karst depression density map for the conterminous United States|
|Authors||Daniel H. Doctor, Jeanne M. Jones, Nathan J. Wood, Jeff T. Falgout, Natalya Igorevna Rapstine|
|Publication Type||Conference Paper|
|Publication Subtype||Conference Paper|
|Record Source||USGS Publications Warehouse|
|USGS Organization||Core Science Analytics and Synthesis; Western Geographic Science Center; Science Analytics and Synthesis; Florence Bascom Geoscience Center|