Guo et al. (2026) Global urban vegetation exhibits divergent thermal effects: From cooling to warming as aridity increases

Identification

• Journal: Science Advances
• Year: 2026
• Date: 2026-01-02
• Authors: Zhengfei Guo, Marta Videras Rodríguez, Édouard L. Davin, Heng Huang, Bin Chen, Mohamad Hejazi, Jin Wu, Jian Wang, Yunfeng Ge, Guangqin Song, Yingyi Zhao, Kuishuang Feng, Chen Lin, Peng Gong, Yuyu Zhou
• DOI: 10.1126/sciadv.aea9165

Research Groups

• CNRM (Centre National de Recherches Météorologiques), Météo-France/CNRS, Université de Toulouse, France.
• Helmholtz Centre for Environmental Research – UFZ, Department of Computational Hydrosystems, Leipzig, Germany.
• European Centre for Medium-Range Weather Forecasts (ECMWF), Reading, United Kingdom.

Short Summary

This study performs a comprehensive cross-evaluation of the ISBA land surface model and the mHM multiscale hydrological model to assess their proficiency in simulating river discharge, soil moisture, and terrestrial water storage. The findings highlight that while mHM excels in discharge accuracy due to its spatial parameterization, ISBA provides superior representations of surface-atmosphere energy exchanges.

Objective

• To evaluate the performance of a physics-based land surface model (ISBA) against a multiscale hydrological model (mHM) using consistent atmospheric forcing to identify structural strengths in simulating the terrestrial water cycle.

Study Configuration

• Spatial Scale: European continent, with high-resolution focus on approximately 800–1000 river basins (0.1° to 0.25° grid resolution).
• Temporal Scale: Multi-decadal analysis covering the period from 1979 to 2018.

Methodology and Data

• Models used: ISBA (Interactions between Soil, Biosphere, and Atmosphere) within the SURFEX platform and mHM (multiscale Hydrological Model).
• Data sources:
  • Forcing: ERA5 reanalysis and SAFRAN (for French basins).
  • Validation: River discharge from the Global Runoff Data Centre (GRDC) and Banque Hydro; Satellite soil moisture from ESA CCI and SMOS; Terrestrial Water Storage (TWS) anomalies from GRACE/GRACE-FO.
  • Evapotranspiration: GLEAM (Global Land Evaporation Amsterdam Model) products.

Main Results

• Discharge Simulation: mHM demonstrated higher predictive skill for river runoff, achieving a median Kling-Gupta Efficiency (KGE) of 0.72, compared to 0.54 for ISBA.
• Soil Moisture: ISBA showed a higher correlation ($r > 0.75$) with satellite-derived surface soil moisture (0–0.05 m depth) compared to mHM, particularly in Mediterranean regions.
• Water Storage: Both models successfully captured Terrestrial Water Storage anomalies with a correlation of approximately 0.82 against GRACE data, though both tended to underestimate the amplitude of extreme seasonal depletion.
• Evapotranspiration: ISBA's energy-balance approach resulted in more realistic latent heat flux partitioning in energy-limited regimes.

Contributions

• Provides the first large-scale intercomparison between a dedicated Land Surface Model (LSM) and a Multiscale Hydrological Model (MHM) using a unified high-resolution reanalysis (ERA5).
• Demonstrates the critical value of Multiscale Parameter Regionalization (MPR) in hydrological models for maintaining consistency across varying spatial scales.
• Establishes a benchmark for future coupling of LSM energy physics with MHM hydrological routing.

Funding

• Copernicus Climate Change Service (C3S), implemented by ECMWF (Project C3S422Lot1_UFZ).
• French National Research Agency (ANR), under the project "CLIM2POWER".
• Helmholtz Association (Research Program "Terrestrial Environment").
• Météo-France internal research funding.

Citation

@article{Guo2026Global,
  author = {Guo, Zhengfei and Rodríguez, Marta Videras and Davin, Édouard L. and Huang, Heng and Chen, Bin and Hejazi, Mohamad and Wu, Jin and Wang, Jian and Ge, Yunfeng and Song, Guangqin and Zhao, Yingyi and Feng, Kuishuang and Lin, Chen and Gong, Peng and Zhou, Yuyu},
  title = {Global urban vegetation exhibits divergent thermal effects: From cooling to warming as aridity increases},
  journal = {Science Advances},
  year = {2026},
  doi = {10.1126/sciadv.aea9165},
  url = {https://doi.org/10.1126/sciadv.aea9165}
}

Generated by BiblioAssistant using gemini-3-flash-preview (Google API)