Enhancing the Optical Trapezoid Model (OPTRAM) for Satellite Remote Sensing of Soil Moisture through Integration of Landcover Information

Abstract

The Optical TRApezoid Model (OPTRAM) has been extensively utilized to map high-resolution surface soil moisture (top 0-5 cm) using surface reflectance observations. OPTRAM parameters, the intercept and slope of the dry and wet edges, are typically obtained by analyzing the data cloud created from the normalized difference vegetation index (NDVI) and the shortwave-infrared transformed reflectance (STR) in the specified region of interest. In this study, we adopt a new approach to calibrate OPTRAM dry and wet edge parameters based on distinctive landcover reflectance properties. In this analysis, we used Sentinel-2 reflectance and the Cropland Data Layer (CDL) landcover datasets via the Google Earth Engine (GEE) to generate 20-m resolution soil moisture maps in Central Valley, California. We evaluated the spatial and temporal accuracy of the original and landcover-specific calibrated OPTRAM against the SMAP-HydroBlocks (HB), a 30-m satellite-based soil moisture dataset, as a well-validated reference. Our results indicate that landcover-specific calibrated OPTRAM significantly improved the accuracy of the soil moisture estimates. The root mean square error (RMSE) was obtained 0.09 m³m⁻³ for the original OPTRAM and 0.05 for the landcover-specific calibrated OPTRAM.