An Improved Approach To Estimate Wind Erosion In Washington’s Columbia Plateau Using Google Earth Engine

Abstract

Loss of soil from erosion leads to declining agricultural productivity, pollution of water bodies, hazardous air quality, and many other detriments to human and environmental health. The agricultural community in Washington State identified soil erosion as a priority concern in the Columbia Plateau. To target specific areas for monitoring and reducing erosion, decision-makers must have access to regional soil loss estimates. The time-consuming, labor-intensive, and site-specific nature of conventional field measurements makes measuring soil erosion at the regional scale nearly impossible. Alternatively, geographic information systems (GIS), remote sensing (RS) datasets, and Google Earth Engine (GEE), a cloud-based geospatial analysis platform, can be used to estimate erosion at the regional scale. In this study, I implemented the Revised Wind Erosion Equation (RWEQ) model to assess wind-driven soil erosion using GEE. Combining a variety of geospatial and RS datasets, I generated monthly erosion estimates for the year 2018 with a spatial resolution of 0.25°. Erosion was lowest during the late fall through winter months, and highest during the summer months when precipitation was scarce and wind speeds were high. In this proof-of-concept study, some key parameters that influence spatiotemporal variability (i.e., tillage, irrigation, plant residue, and wind barriers) were simplified or omitted. The maximum erosion estimate, aggregated to the year, was consistent with results from another study (7.01 and 8.8 kg/m2, respectively). Both values surpass the annual soil loss tolerance threshold established by the US Department of Agriculture (0.25 – 1.25 kg/m2). This range defines the maximum rate of soil loss that can occur while sustaining crop productivity. The estimates from my study emphasize the importance of regional erosion estimates and control measures. My study demonstrates the feasibility of using GIS, RS, and GEE to automate and visualize regional erosion estimates. To encourage transparency, reproducibility, and community collaboration, my code is published in a public GitHub repository. Lastly, my study provides a framework for the future development of a decision-support tool to access and visualize erosion susceptibility in various time periods and regions. Decision-makers can leverage these accessible monthly erosion estimates to guide research and conservation efforts.