LiDAR based hydro-conditioned hydrological modeling for enhancing precise conservation practice placement in agricultural watersheds

Abstract

High resolution Light Detection and Ranging (LiDAR)-derived Digital Elevation Models (DEMs) improve hydrologic modeling and aid in identifying the targeted locations of best conservation practices (CPs) in agricultural watersheds. However, the inability of LiDAR data to represent the conveyance of water under or through the surfaces (i.e., bridges or culverts) impedes the simulated flow, resulting in false upstream pooling. Improper flow simulation affects the accuracy of pollutant load estimations and targeted locations delineated by watershed models or models built upon hydro-conditioned DEMs (hDEM). We propose a novel approach of Hydro-conditioning to modify LiDAR imagery through breach lines, which is essential to accurately replicate the landscape hydrologic connectivity. We compared variations in outcomes of Agricultural Conservation Planning Framework (ACPF), based on manual and automated hDEMs for Plum Creek watershed, Minnesota. The derived flow network, catchment boundaries, drainage areas, locations/number of practices depend on the chosen hDEM. Locations, size and shape of bioreactors, drainage management, farm ponds, nutrient removal wetlands, riparian buffers are severely affected by hydro-conditioning. Shuttle Radar Topography Mission (SRTM) validation of hDEMs showed that Mean Average Percentage Deviation (MAPE) for automated and manual hDEMs is 1.34 and 0.998 respectively. Also, proximity analysis with a buffer of 200 m showed that CPs’ locations delineated by manual hDEM match better with the existing ones as compared to automated hDEM. Results indicate that coupled approach of using automated and manual ‘hDEM’ is best suited for guiding stakeholders towards the field-scale planning in a cost-saving manner.