A Mechanism-Learning Deeply Coupled Model for Remote Sensing Retrieval of Global Land Surface Temperature

2025-04-10

AI Summary

Single-band land surface temperature (LST) retrieval from remote sensing is an ill-posed inverse problem, suffering from severe accuracy degradation under data scarcity and high-humidity conditions. Method: This paper proposes a mechanism-driven, physics-informed deep learning paradigm: embedding the radiative transfer equation (RTE) directly into a neural network architecture to construct a physically constrained, coupled inversion framework. The approach integrates atmospheric profile–driven synthetic data generation, physics-informed neural networks (PINNs), and multi-constraint joint optimization. Contribution/Results: Experiments demonstrate a 30% reduction in global root-mean-square error (RMSE); under extreme humidity, mean absolute error (MAE) improves from 4.87 K to 2.29 K (53% reduction). Continental-scale validation across five continents confirms strong generalizability. This work achieves, for the first time, end-to-end differentiable coupling of first-principles radiative transfer physics with deep learning, overcoming long-standing limitations in physical consistency and robustness of single-channel LST retrieval.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Low Level & Physics-based Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Land surface temperature (LST) retrieval from remote sensing data is pivotal for analyzing climate processes and surface energy budgets. However, LST retrieval is an ill-posed inverse problem, which becomes particularly severe when only a single band is available. In this paper, we propose a deeply coupled framework integrating mechanistic modeling and machine learning to enhance the accuracy and generalizability of single-channel LST retrieval. Training samples are generated using a physically-based radiative transfer model and a global collection of 5810 atmospheric profiles. A physics-informed machine learning framework is proposed to systematically incorporate the first principles from classical physical inversion models into the learning workflow, with optimization constrained by radiative transfer equations. Global validation demonstrated a 30% reduction in root-mean-square error versus standalone methods. Under extreme humidity, the mean absolute error decreased from 4.87 K to 2.29 K (53% improvement). Continental-scale tests across five continents confirmed the superior generalizability of this model.

Problem

• Improving single-channel land surface temperature retrieval accuracy
• Combining mechanistic modeling with machine learning for LST
• Reducing errors in global LST retrieval under extreme conditions

Innovation

• Deeply coupled mechanistic and machine learning model
• Physics-informed framework with radiative transfer constraints
• Global validation showing 30% error reduction