CHAM-net: A Contrastive Hierarchical Adaptive Meta-network for Robust Global Methane Flux Prediction
Year: 2026
Type: conference-paper
Abstract: Methane is a potent greenhouse gas that significantly contributes to global warming. However, accurately estimating global methane emissions and consumption remains challenging due to the complex interactions among environmental drivers that may vary across spatial and temporal scales. Prior data-driven methods often overlook the inherent spatiotemporal heterogeneity of ecosystems, failing to explicitly capture site-specific characteristics and cross-year evolutionary dynamics. To address these issues, we propose the Contrastive Hierarchical Adaptive Meta-network (CHAM-net), a novel framework that explicitly learns from historical context to capture site-specific dynamics. CHAM-net employs a hierarchical encoder–decoder architecture, in which the encoder captures site-specific characteristics from historical data and then dynamically conditions the decoder to generate the final prediction. Experimental results demonstrate that CHAM-net consistently outperforms all baseline methods on both simulation and observational datasets for methane emission and consumption, achieving nRMSE values as low as 0.43 and 0.88 with corresponding R^2 scores up to 0.97 and 0.68 for emission prediction.
Language: English
Subfield: Global and Planetary Change
Field: Environmental Science
Domain: Physical Sciences
SDG: Climate action
Open Access status: bronze
Awards: Collaborative Research: CAIG: Toward Next-Generation Global Forest Carbon Monitoring via Integrated Sensing, Modeling and AI to Advance Carbon Cycle Science, Collaborative Research: CAIG: Toward Next-Generation Global Forest Carbon Monitoring via Integrated Sensing, Modeling and AI to Advance Carbon Cycle Science, G2030118026, CDS&E: Physics Guided Super-Resolution for Turbulent Transport, FAI: Advancing Deep Learning Towards Spatial Fairness, Full Proposal: Environmental Data Science Innovation and Impact Lab (ESIIL): Accelerating Discovery by Fostering an Open and Diverse Earth Data Revolution, Collaborative Research: CAIG: Emulating Water Isotopes in Fully-coupled Global Climate Models using Knowledge-guided Machine Learning, Collaborative Research: III: Small: Physics Guided Graph Networks for Modeling Water Dynamics in Freshwater Ecosystems, G4552795344, CAREER: Combining Machine Learning and Physics-based Modeling Approaches for Accelerating Scientific Discovery, Collaborative Research: CAIG: Emulating Water Isotopes in Fully-coupled Global Climate Models using Knowledge-guided Machine Learning, G5334123415, G6982498602, Collaborative Research: EarthCube Capabilities: ICESpark: An Open-Source Big Data Platform for Science Discoveries in the New Arctic and Beyond
OpenAlex ID: W7163197873