[A-HW27] Biodiversity, nutrients and other materials in ecosystems from headwaters to coasts

[AHW27-01] Development of a Model Framework for Terrestrial Carbon Flux Prediction

★Invited Papers

*Adina Paytan¹, Ashley Brereton¹ (1.University of California Santa Cruz)

Keywords:Machine Learning, Wetlands, Greenhouse Gases, Upscalling

Wetlands play a pivotal role in carbon sequestration but emit methane (CH4), creating uncertainty in their net climate impact. Although process-based models offer mechanistic insights into wetland dynamics, they are computationally expensive, uncertain, and difficult to upscale. In contrast, data-driven models provide a scalable alternative by leveraging extensive datasets to identify patterns and relationships, making them more adaptable for large-scale applications under current climate conditions. However, their performance can vary significantly depending on the quality and representativeness of the data, as well as the model design, raising questions about their reliability and generalizability in diverse wetland contexts. To address these issues, we present a data-driven framework for upscaling wetland CO2 and CH4 emissions, across a range of machine learning models that vary in complexity, validated against observational data from the Sacramento-San Joaquin Delta. We show that deep learning approaches outperform linear regression models. However, interannual variability is less well captured. By integrating vertically-resolved atmospheric, subsurface, and spectral reflectance information from readily available sources, the model identifies key drivers of wetland CO2 and CH4 emissions and enables regional upscaling. These findings demonstrate the potential of data-driven methods for upscaling, providing practical tools for wetland management and restoration planning to support climate mitigation efforts.