A New Analytic Framework and Notebook for Terrain Analysis

Terrain analysis algorithms evolved along the same timeline as digital elevation model (DEM) spatial resolutions, with the typical implementation using only immediately adjacent cells (3x3 window) being appropriate for the landscape processes of interest. Finer resolution DEMs require new approaches to effectively model terrain forms and processes. Here, we present a new approach based on recent innovations and some longstanding ideas in terrain analysis, that allow users to control the scale of analysis through a window size, use improved polynomial surface fitting, and utilize geodesic calculations. This new framework has been implemented in ArcGIS to calculate aspect, slope, and multiple curvature types. To share our implementation methodology and enable others to build from it we released a Python implementation of the tool in a modular, tutorial style as a Jupyter Notebook. Researchers and analysts can use this notebook to further experiment with the tool parameters and implement additional terrain metrics of interest.