MetaFlux: Meta-learning global carbon fluxes from sparse spatiotemporal observations

Abstract

We provide a global, long-term carbon flux dataset of gross primary production and ecosystem respiration generated using meta-learning, called MetaFlux. The idea behind meta-learning stems from the need to learn efficiently given sparse data by learning how to learn broad features across tasks to better infer other poorly sampled ones. Using meta-trained ensemble of deep models, we generate global carbon products on daily and monthly timescales at a 0.25-degree spatial resolution from 2001 to 2021, through a combination of reanalysis and remote-sensing products. Site-level validation finds that MetaFlux ensembles have lower validation error by 5–7% compared to their non-meta-trained counterparts. In addition, they are more robust to extreme observations, with 4–24% lower errors. We also checked for seasonality, interannual variability, and correlation to solar-induced fluorescence of the upscaled product and found that MetaFlux outperformed other machine-learning based carbon product, especially in the tropics and semi-arids by 10–40%. Overall, MetaFlux can be used to study a wide range of biogeochemical processes.

Improved estimation of lake greenhouse gas emissions in China based on nutrients and potential benefits from eutrophication controls

“…These advanced data sources contribute to overcoming the uncertainties associated with data sparsity and provide a more precise understanding of regional-scale emissions 55 . Furthermore, the development of machine learning models, including meta-learning approaches, can help address data gaps and enhance the quality of estimates 56 . These models have the capability to fill in missing data and improve the accuracy of GHGs emissions estimations from lakes.…”

Exploring the Potential of Long Short‐Term Memory Networks for Predicting Net CO2 Exchange Across Various Ecosystems With Multi‐Source Data

“…The Long Short‐Term Memory model (LSTM) is a dynamic statistical method that has demonstrated excellent performance on sequence data, such as crop field classification (Rußwurm & Körner, 2018). With its distinctive design, the LSTM model can effectively address long‐term considerations and incorporate memory effects of climate and vegetation, thus aiding in the representation of interannual fluctuations in carbon fluxes (Besnard et al., 2019; Nathaniel et al., 2023). To support this concept, we developed and applied an LSTM model to predict site‐level NEE in North America.…”