An automated method to extract typical karst landform entities from contour lines on topographic maps

Mapping specific landform entities in the past is mainly achieved by manually examining contour lines on topographic maps. Automated delimitation of specific landform from digital elevation data remains difficult in geomorphologic mapping. This paper presents a method to automatically identify five common surface features of karst landscapes: isolated karst hill or sinkhole, clustered karst hills or sinkholes, and clustered hills with sinkholes. These landform entities have their own singular geomorphometric characteristics and thus could be identified from digital elevation data. In this study, boundary of individual landform entity is defined by an outmost closed contour line (CCL), which contains at least another CCL but not contained by any other CCLs. The innermost CCL with a local extreme elevation value represents either a peak or a sink of that specific landform. Between the innermost and the outmost CCLs, several intermediate CCLs may exist, depending upon relief of the landform and contour interval of the topographic maps. The aforementioned surface karst landform entities were then delimitated and identified by examining the spatial relationships among these CCLs and the change of their elevation values. The method was applied to delimitate these landform entities in Oolitic, IN and Florida, PR. Comparison of the distribution of these surface features in these two areas provides new insights into karst development processes and landscape evolution.