UAV applications in humid tropical fluvial geomorphology 

Abstract

<p>The use of unmanned aerial vehicles (UAVs or “drones”) in geosciences is growing rapidly and is significantly contributing to the geomorphological analysis of diverse landforms and their dynamics in a variety of environments around the world. Specifically, UAVs in geomorphology have proved to be a robust and versatile tool due to their ability to acquire large volumes of high-resolution data in a relatively short time and at low-cost compared to other surveying methods. This technology is evolving rapidly and is becoming each time more relevant for the analysis of surface processes. It allows the rapid mapping of highly dynamic geomorphological features with high-resolution which is particularly useful for the study of fluvial processes. Depending on the environmental conditions and flight parameters, common low-cost (<$3000) commercial UAV models such as the DJI Phantom 4 V.2 used in this research can map over 30 hectares of surface in less than 20 minutes, proving to be a valuable tool for monitoring fluvial processes, especially in tropical humid environments where cloudiness can be an important obstacle for other survey methods such as photogrammetric-piloted airborne missions or satellite imagery. We present the results of the usefulness of UAV technology in fluvial geomorphology with several case studies in diverse humid tropical environments of Costa Rica, Central America.  We illustrate how UAV surveying can acquire high-resolution digital terrain and surface models for hydraulic modeling, geomorphic analysis, and forensic reconstruction of flood events in two urban rivers. We also illustrate how low-altitude photogrammetric flights can detect channel changes with centimetric resolution under dense pristine forest coverage, a challenging environment for monitoring river dynamics. Additionally, we used UAV photogrammetric post-processing for automated analysis of particle size distribution detection in river channels. Finally, we exemplify how this technology can be used for riparian vegetation analysis under diverse conditions. The post-processing of the orthomosaics and surface models allowed to construct high-resolution indexes such as Visible Atmospherically Resistant Index (VARI), vegetation height, Leaf Area Index (LAI), and other datasets for ecological analysis of riverine environments.</p>