Seasonal Change Detection of Riverscape Terrain and Vegetation using Drone Laser Scanning

Riverscapes are complex ecosystems consisting of interconnected components including stream channels, banks, riparian zones, and floodplains. Physical measures of these environments, such as channel geometry, floodplain connectivity, and vegetative cover, are important for a range of hydrological and ecological processes. Many of these processes occur at very fine spatial scales and can fluctuate dramatically over time. As a result, the spatial and temporal complexity of a riverscape is difficult to quantify. In this study, we used drone laser scanning (DLS), or drone-based lidar, as a high-resolution surveying tool for measuring riverscape terrain and vegetation. DLS combines the precision of lidar with the versatility of unmanned aerial vehicles. We scanned a 0.7-km reach of Stroubles Creek, located in Blacksburg, VA, five times over eighteen months (April, August, and November of 2017 and April and October of 2018) to allow for seasonal comparisons. The DLS surveys had an average point spacing of 5 cm or 400 points/m². We developed data pipelines utilizing ArcGIS and LAStools to classify the raw lidar data into ground and vegetation points. Once classified, we created high-resolution raster models representing different aspects of the riverscape environment, such as terrain elevation, canopy height, and roughness. We compared the resulting models and calculated DEMs of difference to provide a quantitative measure of seasonal change in terrain (representing migration of the stream channel, erosion, deposition, etc.) and vegetation (representing change in canopy height, vegetative cover, etc.). DLS is unique in its ability to measure both the terrain and vegetation over the entire riverscape from the stream channel to the floodplain at fine spatial and temporal resolutions not possible with lidar from planes or helicopters.