A fully automated OPTRAM (aOPTRAM) for soil moisture retrieval: Evaluating multiple fitting functions, vegetation indices, land-cover types, and scales

Soil moisture is a key variable in hydrological and climate studies. The Optical Trapezoid Model (OPTRAM) enables high-resolution soil moisture retrieval using only reflective bands, but most implementations still rely on manual edge/parameter selection and have not been systematically evaluated across diverse ecosystems. Here, we integrate an extended Adaptive Sliding Window (ASW) edge-detection algorithm into an automated OPTRAM (aOPTRAM) workflow to assess its performance using Sentinel-2 imagery and in-situ measurements from 10 networks (130 stations) spanning bare soil, shrubland, grassland, and tree cover. We compared vegetation indices (NDVI, SAVI), edge-fitting functions (linear, exponential, polynomial), and spatial scales (network vs. station), and evaluated accuracy using Pearson's correlation coefficient (R), unbiased root-mean-square error (ubRMSE), bias, and Kling-Gupta efficiency (KGE). aOPTRAM performance is sensitive to vegetation index, land cover, edge functional form, and export scale. SAVI with linear edges performs best in shrublands, whereas NDVI with exponential edges is generally preferred in grasslands and tree cover. Bare-soil remains challenging because the STR dynamic range is small, leading to weak wet–dry separability and amplified retrieval uncertainty. Station-scale export consistently improves accuracy relative to the network scale and better captures seasonal soil-moisture dynamics. Overall, aOPTRAM achieves performance comparable to optimal OPTRAM while providing a fast, edge-calibration-free framework for high-resolution soil moisture monitoring across heterogeneous landscapes.