A Physics-Based Feature XGBoost Model for Frozen Ground Surface Reflectance Reconstruction on the Tibetan Plateau (opens in new tab)

Surface reflectance (SR) variability in frozen ground regions is critical for understanding land surface processes and their response to climate change. However, traditional physical models rely on complex input parameters, limiting their adaptability across diverse environments, while statistical and machine learning approaches often neglect the underlying physical mechanisms, resulting in constrained reconstruction accuracy. To address these limitations, we propose a physics-based feature X...