Huang et al. (2026) Improved estimation of net ecosystem CO 2 exchange over North America using LSTM-based flux upscaling (2001–2021)

Identification

• Journal: Earth system science data
• Year: 2026
• Date: 2026-10-07
• Authors: Chengcheng Huang, Wei He, Brendan Byrne, Ngoc Tu Nguyen, Jingfeng Xiao, Hua Yang, Philippe Ciais, Songhan Wang, Xing Li, Han Ma, Peipei Xu, Mengyao Zhao, Hui Chen, Weimin Ju
• DOI: 10.5194/essd-18-7367-2026

Research Groups

• State Key Laboratory of Earth Surface Processes and Disaster Risk Reduction, Faculty of Geographical Science, Beijing Normal University, China
• Zhejiang Carbon Neutral Innovation Institute & Zhejiang International Cooperation Base for Science and Technology on Carbon Emission Reduction and Monitoring, Zhejiang University of Technology, China
• International Institute for Earth System Science, Nanjing University, China
• Jet Propulsion Laboratory, California Institute of Technology, USA
• Institute of Remote Sensing and Earth Sciences, Hangzhou Normal University, China
• Earth Systems Research Center, Institute for the Study of Earth, Oceans, and Space, University of New Hampshire, USA
• Laboratoire des Sciences du Climat et de l'Environnement, LSCE/IPSL, CEA-CNRS-UVSQ, Université Paris-Saclay, France
• Jiangsu Collaborative Innovation Center for Modern Crop Production, Key Laboratory of Crop Physiology and Ecology in Southern China, Nanjing Agricultural University, China
• School of Geography and Planning, Sun Yat-sen University, Guangzhou, China
• JC STEM Lab of Quantitative Remote Sensing, Department of Geography, The University of Hong Kong, Hong Kong
• International Center for China Development Studies, The University of Hong Kong, Hong Kong
• School of Geography and Tourism, Anhui Normal University, Wuhu, Anhui, China
• International Institute for Earth System Science, Nanjing University, Jiangsu, China

Short Summary

This study presents an improved estimation of net ecosystem CO2 exchange (NEE) over North America using a Long Short-Term Memory (LSTM)-based flux upscaling method. The resulting dataset, named MemoryFlux, provides monthly NEE estimates at 0.1° spatial resolution from 2001 to 2021.

Objective

• Investigate the effectiveness of LSTM-based flux upscaling in estimating regional-scale terrestrial carbon budgets.
• Evaluate the consistency and differences between MemoryFlux and existing bottom-up products, as well as independent top-down inversion estimates.

Study Configuration

• Spatial Scale: North America, at a spatial resolution of 0.1°.
• Temporal Scale: Monthly intervals from 2001 to 2021.

Methodology and Data

• Models used: Long Short-Term Memory (LSTM) network.
• Data sources:
  • CO2 eddy covariance flux data from FLUXNET2015 and AmeriFlux networks.
  • Climate reanalysis data from ERA5-Land.
  • Satellite remote sensing datasets, including NDVI, LAI, FAPAR, SIF, and MODIS land cover.

Main Results

• MemoryFlux captures the spatial patterns of NEE over North America, with strong carbon uptake in the southeastern United States and major carbon sources in northern and central Canada.
• The dataset shows greater consistency with independent atmospheric inversion estimates than existing bottom-up upscaling products.
• MemoryFlux represents regional-scale seasonal, interannual, and extreme-event-related variability in NEE.

Contributions

• This study provides an improved bottom-up estimation of North American NEE using LSTM-based flux upscaling.
• The resulting dataset, MemoryFlux, serves as a valuable resource for investigating terrestrial carbon dynamics and ecosystem-climate interactions.

Funding

• Not specified.

Citation

@article{Huang2026Improved,
  author = {Huang, Chengcheng and He, Wei and Byrne, Brendan Keith Aidan and Nguyen, Ngoc Tu and Xiao, Jingfeng and Yang, Hua and Ciais, Philippe and Wang, Songhan and Li, Xing and Ma, Han and Xu, Peipei and Zhao, Mengyao and Chen, Hui and Ju, Weimin},
  title = {Improved estimation of net ecosystem CO 2 exchange over North America using LSTM-based flux upscaling (2001–2021)},
  journal = {Earth system science data},
  year = {2026},
  doi = {10.5194/essd-18-7367-2026},
  url = {https://doi.org/10.5194/essd-18-7367-2026}
}