High-Resolution Global Land Surface Temperature Retrieval via a Coupled Mechanism-Machine Learning Framework

2025-09-05

AI Summary

Land surface temperature (LST) retrieval from remote sensing suffers from low accuracy over heterogeneous surfaces and under extreme atmospheric conditions—particularly hot-humid environments—due to large biases in conventional split-window algorithms and poor interpretability and generalizability of purely data-driven machine learning approaches. Method: This study proposes a tightly coupled inversion framework integrating physical mechanism–driven and data-driven paradigms: it synergistically embeds radiative transfer modeling, MODTRAN-based atmospheric simulation, and physics-constrained optimization, augmented by multi-source remote sensing observations and global atmospheric profile datasets. Contribution/Results: The framework achieves both high physical interpretability and strong nonlinear modeling capability. Validated across 29 global sites, it attains MAE = 1.84 K, RMSE = 2.55 K, and R² = 0.966; under extreme conditions, retrieval errors decrease by over 50%. It significantly enhances LST retrieval stability, cross-regional generalizability, and reliability for climate and ecological applications.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Land surface temperature (LST) is vital for land-atmosphere interactions and climate processes. Accurate LST retrieval remains challenging under heterogeneous land cover and extreme atmospheric conditions. Traditional split window (SW) algorithms show biases in humid environments; purely machine learning (ML) methods lack interpretability and generalize poorly with limited data. We propose a coupled mechanism model-ML (MM-ML) framework integrating physical constraints with data-driven learning for robust LST retrieval. Our approach fuses radiative transfer modeling with data components, uses MODTRAN simulations with global atmospheric profiles, and employs physics-constrained optimization. Validation against 4,450 observations from 29 global sites shows MM-ML achieves MAE=1.84K, RMSE=2.55K, and R-squared=0.966, outperforming conventional methods. Under extreme conditions, MM-ML reduces errors by over 50%. Sensitivity analysis indicates LST estimates are most sensitive to sensor radiance, then water vapor, and less to emissivity, with MM-ML showing superior stability. These results demonstrate the effectiveness of our coupled modeling strategy for retrieving geophysical parameters. The MM-ML framework combines physical interpretability with nonlinear modeling capacity, enabling reliable LST retrieval in complex environments and supporting climate monitoring and ecosystem studies.

Problem

• Retrieving accurate land surface temperature under heterogeneous conditions
• Overcoming biases in humid environments with traditional algorithms
• Improving interpretability and generalization of machine learning methods

Innovation

• Coupled mechanism-ML framework for LST retrieval
• Integrates radiative transfer with data-driven learning
• Physics-constrained optimization using MODTRAN simulations