Luo et al. (2026) Hydroclimate shapes photosynthetic sensitivity to cloud cover across global terrestrial ecosystems

Identification

• Journal: Nature Communications
• Year: 2026
• Date: 2026-02-12
• Authors: Hao Luo, Ana Bastos, Markus Reichstein, Gregory Duveiller, Jan Kretzschmar, Johannes Quaas
• DOI: 10.1038/s41467-026-69480-3

Research Groups

• Leipzig Institute for Meteorology, Leipzig University, Leipzig, Germany
• Institute for Earth System Science and Remote Sensing, Leipzig University, Leipzig, Germany
• Max Planck Institute for Biogeochemistry, Jena, Germany
• German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Leipzig, Germany

Short Summary

This study reveals that the sensitivity of terrestrial photosynthesis to cloud cover is spatially determined by hydroclimate, with clouds promoting photosynthesis in water-limited arid regions (via precipitation) and inhibiting it in energy-limited humid regions (via sunlight blockage). Under future warming, this leads to projected declines in gross primary productivity in arid regions and increases in humid regions, exacerbating regional disparities.

Objective

• To quantify how the sensitivity of global terrestrial photosynthesis to cloud cover is shaped by hydroclimate, distinguishing between cloud-mediated radiative and precipitative effects, and to project how future cloud cover changes under global warming will impact vegetation productivity.

Study Configuration

• Spatial Scale: Global terrestrial ecosystems, ranging from site-level (FLUXNET) to global gridded datasets (0.05° × 0.05° to 2° × 2° resolution, often aggregated to 1° × 1°).
• Temporal Scale: Decades (e.g., 1984–2099), with analyses at daily, 8-daily, monthly, and annual resolutions. Delayed effects of precipitation are typically within 1 month.

Methodology and Data

• Models used:
  • TRENDYv12 project (ensemble of 20 dynamic global vegetation models)
  • FLUXCOM initiative (machine learning-based global flux products)
  • Light Use Efficiency (LUE) model (for MODIS GPP)
  • Penman–Monteith equation (for potential evapotranspiration, PET)
  • Radiative transfer modeling (for CERES photosynthetically active radiation, PAR)
• Data sources:
  • In-situ: FLUXNET2015 eddy covariance measurements (Gross Primary Productivity (GPP), precipitation amount (PA), surface incoming shortwave radiation (SWin), air temperature, wind speed, latent heat flux, sensible heat flux, vapor pressure deficit (VPD)).
  • Satellite:
    • MODIS/Terra: Cloud Fraction (CF), single-layer liquid/ice cloud fraction (CFliquid, CFice), Cloud Optical Thickness (COT), joint histograms of COT and Cloud Top Pressure (CTP), MOD17A2HGF GPP, MCD12C1 land cover product.
    • OCO-2: Discrete Solar-Induced Fluorescence (SIF) retrievals (for GOSIF).
    • AMSR-E, AMSR2: X-band Vegetation Optical Depth (VOD) from LPDR v3.
    • CERES: Surface downwelling solar radiation, PAR.
    • GPM IMERG: Precipitation amount (PA).
    • ISCCP-HGM: Global monthly cloud parameters.
  • Reanalysis:
    • ERA5: Cloud Fraction (CF), meteorological parameters (for FLUXCOM).
    • MERRA-2: Meteorological reanalysis (for GOSIF).
    • CRU JRA v2.5: 6-hourly meteorological reanalysis (for TRENDY and annual scale GPP-to-CF sensitivity).
  • Gridded Observations/Reanalysis:
    • FLUXCOM-RS, FLUXCOM-RSMETEOERA5, FLUXCOM-X-BASE: GPP datasets.
    • GOSIF: Solar-Induced Fluorescence (SIF).
    • CRU TS v4.08: Monthly PA and PET (for Humidity Index), annual average CF, 2 meter temperature (T2m).
  • Climate Models: CMIP6 (ensemble mean of 39 models for historical and ssp585 future CF projections).

Main Results

• The sensitivity of terrestrial photosynthesis to cloud cover is spatially determined by hydroclimate, quantified by the humidity index (mean annual precipitation-to-evapotranspiration ratio).
• In water-limited arid regions (low humidity index), clouds promote photosynthesis primarily through increased precipitation, with a delayed effect typically within 1 month.
• In energy-limited humid regions (high humidity index), clouds inhibit photosynthesis almost instantaneously by blocking sunlight.
• This spatial pattern is robustly observed across site-level eddy covariance measurements (FLUXNET) and global gridded datasets (FLUXCOM-RS), and validated by offline dynamic global vegetation models (TRENDYv12) and time-lagged causal analysis.
• The sensitivity of GPP to cloud fraction is effectively explained by the balance between its sensitivity to precipitation and radiation.
• Liquid clouds exert a stronger influence on GPP sensitivity than ice clouds, and GPP sensitivity to grid-mean Cloud Optical Thickness (COT) shows similar hydroclimate-shaped spatial patterns.
• Under a warming climate (CMIP6 ssp585 scenario, 2015–2099), global average cloud fraction is projected to decline.
• This decline in cloud cover is projected to cause GPP to decrease in arid regions and increase in humid regions, leading to a spatial shift in GPP from arid to humid areas and exacerbating regional disparities in ecosystem carbon sequestration capacity.
• Cloud fraction serves as an effective combined proxy for precipitation and radiation, performing similarly to a multiple regression model using both variables to predict GPP.

Contributions

• Provides the first comprehensive global assessment of the net impact of cloud-induced changes in surface solar radiation and precipitation on terrestrial photosynthesis.
• Identifies the humidity index as the primary hydroclimatic factor shaping the spatial patterns of photosynthetic sensitivity to cloud cover.
• Quantifies the distinct temporal responses of photosynthesis to clouds in different hydroclimates: instantaneous inhibition in humid regions versus delayed promotion in arid regions.
• Projects future shifts in global GPP patterns (decline in arid, increase in humid regions) driven by cloud cover changes under a strong warming scenario (CMIP6 ssp585), highlighting potential exacerbation of regional ecosystem function disparities.
• Demonstrates the robustness and utility of using total cloud fraction as a single, effective proxy for the combined effects of cloud-mediated radiation and precipitation on GPP.
• Employs a multi-dataset, multi-scale approach (in-situ, satellite, reanalysis, and dynamic global vegetation models) to ensure the reliability and causality of the findings.

Funding

• European Research Council (ERC) Synergy Grant “understanding and modeling the Earth system with machine learning (USMILE)” (grant agreement no. 855187)
• European Union’s Horizon 2020 research and innovation program
• Projekt DEAL (Open Access funding)

Citation

@article{Luo2026Hydroclimate,
  author = {Luo, Hao and Bastos, Ana and Reichstein, Markus and Duveiller, Gregory and Kretzschmar, Jan and Quaas, Johannes},
  title = {Hydroclimate shapes photosynthetic sensitivity to cloud cover across global terrestrial ecosystems},
  journal = {Nature Communications},
  year = {2026},
  doi = {10.1038/s41467-026-69480-3},
  url = {https://doi.org/10.1038/s41467-026-69480-3}
}