Indicators of wind-driven forest disturbances – pit–mound topography, its automatic detection and significance

Abstract

Pit-and-mound topography is a result of tree uprooting caused by hurricane-force wind events and hence can act as a bioindicator of forest disturbance. The occurrence and evolution of pit–mound topography can be analyzed using detailed elevation data, such as point clouds from Light Detection and Ranging (LiDAR) surveys. The objective of this study was to develop an automatic method of pit–mound topography detection. We propose the usage of closed contour lines to extract the locations of pits and mounds. We performed analyses in two study areas (Markowa and Stonów) located on the Babia Góra Massif (southern Poland). We computed the digital elevation model (DEM), extracted contour lines, calculated the length of each contour line and selected only closed contours belonging to a specified length interval. Then, we created polygons from the outermost closed contour lines. We classified polygons into “pits” and “mounds” by investigating the location of the highest and lowest altitudes within the polygon. We tested 27 variants of our method using different DEM spatial resolutions, contour intervals and contour length intervals. To estimate the accuracy of our method, we created a validation dataset by performing manual recognition of pit–mound pairs based on the topographic position index (TPI). One of the highest accuracies, obtained for the 1st variant of our method, reached 96.9 % for pits, 93.8 % for mounds and 90.6 % for pit–mound pairs in the Stonów area. In the Markowa area, this variant achieved an accuracy of 95.2 % for pits, 90.5 % for mounds and 85.7 % for pit–mound pairs. Our method can be used as an important step in analyses conducted in forest ecology, geomorphology or soil science.