Zhou et al. (2025) Neglecting land–atmosphere feedbacks overestimates climate-driven increases in evapotranspiration

Identification

• Journal: Nature Climate Change
• Year: 2025
• Date: 2025-09-11
• Authors: Sha Zhou, Bofu Yu
• DOI: 10.1038/s41558-025-02428-5

Research Groups

• State Key Laboratory of Earth Surface Processes and Disaster Risk Reduction, Faculty of Geographical Science, Beijing Normal University, Beijing, China
• Institute of Land Surface System and Sustainable Development, Faculty of Geographical Science, Beijing Normal University, Beijing, China
• School of Engineering and Built Environment, Griffith University Nathan Campus, Brisbane, Queensland, Australia

Short Summary

This study develops a theoretical framework to disentangle land-atmosphere interactions, achieving consistent evapotranspiration (ET) projections between offline and coupled models. It reveals that neglecting these feedbacks leads to a 25–39% overestimation of climate-driven global ET increases and a 77–121% exaggeration of negative land surface contributions, causing significant discrepancies in hydrological projections.

Objective

• To develop a theoretical framework that disentangles land-atmosphere interactions to achieve highly consistent evapotranspiration (ET) projections between offline and coupled models, thereby improving the understanding of hydrological responses to climate warming.

Study Configuration

• Spatial Scale: Global terrestrial and oceanic grid cells.
• Temporal Scale: Historical (e.g., 1980-2009), future (e.g., 2071-2100), and 140-year simulations (1850-1989) with 30-year averaging periods.

Methodology and Data

• Models used: Coupled Model Intercomparison Project Phase 6 (CMIP6) models (32 models for some analyses, 7 Earth System Models for others), Budyko framework.
• Data sources: Fluxnet2015 dataset, CMIP6 model simulations (publicly available at https://esgf-node.llnl.gov/search/cmip6/), Natural Earth basemap.

Main Results

• A theoretical framework was developed that successfully disentangles land-atmosphere interactions, leading to highly consistent evapotranspiration (ET) projections between offline and coupled models.
• Previous estimates of climate-driven ET increases were found to be exaggerated due to a substantial overestimation of atmospheric evaporative demand.
• Atmospheric conditions often assumed to drive ET are, in fact, responses to ET changes induced by soil moisture and vegetation dynamics.
• Neglecting land-atmosphere feedbacks resulted in a 25–39% overestimation of climate-driven global ET increases.
• This neglect also led to a 77–121% exaggeration of the negative contribution from land surface changes to ET.
• These biases are identified as the primary cause of large discrepancies in hydrological projections and attributions between offline and coupled models.

Contributions

• Development of a novel theoretical framework that effectively disentangles complex land-atmosphere interactions, enabling more consistent and reliable ET projections.
• Quantification of the significant biases (25–39% global ET overestimation, 77–121% land surface contribution exaggeration) in climate-driven hydrological projections caused by neglecting land-atmosphere feedbacks.
• Reconciliation of discrepancies in hydrological projections and attributions between offline and coupled models, enhancing the robustness of future hydrological assessments.
• Underscoring the critical importance of accurately representing land-atmosphere interactions for improving the reliability and consistency of future hydrological projections in a warming climate.

Funding

• National Natural Science Foundation of China (grant nos. 42471108 and 42521001)
• National Key Research and Development Program of China (grant no. 2022YFF0801303)

Citation

@article{Zhou2025Neglecting,
  author = {Zhou, Sha and Yu, Bofu},
  title = {Neglecting land–atmosphere feedbacks overestimates climate-driven increases in evapotranspiration},
  journal = {Nature Climate Change},
  year = {2025},
  doi = {10.1038/s41558-025-02428-5},
  url = {https://doi.org/10.1038/s41558-025-02428-5}
}