Ayala et al. (2026) Integration of the Global Water and Lake Sectors within the ISIMIP framework through scaling of streamflow inputs to lakes

Identification

• Journal: Geoscientific model development
• Year: 2026
• Date: 2026-01-05
• Authors: Ana I. Ayala, José L. Hinostroza, Daniel Mercado-Bettin, Rafael Marcé, Simon N. Gosling, Donald C. Pierson, Sebastian Sobek
• DOI: 10.5194/gmd-19-41-2026

Research Groups

• Limnology Unit, Department of Ecology and Genetics, Uppsala University, Uppsala, Sweden
• Faculty of Civil Engineering, National University of Engineering, Lima, Perú
• Centre for Advanced Studies, National Spanish Research Council (CEAB-CSIC), Blanes, Spain
• School of Geography, University of Nottingham, Nottingham, United Kingdom

Short Summary

This study developed and validated a novel scaling methodology to dynamically link gridded hydrological model outputs from the ISIMIP Global Water Sector to individual lake catchments in the ISIMIP Lake Sector, demonstrating its reliability for estimating lake streamflow inputs.

Objective

• To establish the first dynamic connection between the ISIMIP Global Water and Lake Sectors by developing a scaling methodology to estimate streamflow inputs to individual lakes from gridded hydrological model outputs, thereby enabling more comprehensive climate change impact assessments on lakes.

Study Configuration

• Spatial Scale: Global (0.5° by 0.5° grid resolution for water fluxes), with a focus on 70 representative lakes across Sweden. Lake surface areas ranged from 0.34 km² to 5486 km², and catchment areas from 101 km² to 48 421 km².
• Temporal Scale: WaterGAP 2.2e simulations from 1901 to 2019 (monthly outputs). Validation against HYPE model simulations and observed data for the period 1981–2010 (monthly and annual averages).

Methodology and Data

• Models used:
  • WaterGAP 2.2e (Water Global Assessment and Prognosis): A process-based global hydrological model.
  • HYPE (Hydrological Predictions for the Environment): A hydrological model used for validation, simulating processes at the catchment scale across Europe.
• Data sources:
  • ISIMIP3a Global Water Sector outputs (from WaterGAP 2.2e): Monthly total runoff (qtot in m s⁻¹), groundwater runoff (qg in m s⁻¹), and discharge (dis in m³ s⁻¹).
  • ISIMIP3a Global Lake Sector: Representative lakes and their catchment areas sourced from the HydroLAKES database.
  • GSWP3-W5E5 climate forcing dataset (obsclim) and direct human forcing (histsoc) for WaterGAP 2.2e.
  • ERA5 reanalysis climate data for forcing the HYPE model.
  • Observed daily river discharge data from the Swedish Meteorological and Hydrology Agency (SMHI) for 10 lakes.

Main Results

• The developed scaling approach reliably estimated streamflow to individual lakes, showing good agreement with both reference (HYPE model) and observed data across various hydrological settings.
• For monthly streamflow (1981–2010) across 70 sites, the average Kling-Gupta efficiency (KGE) against HYPE reference data was 0.59 ± 0.18 (mean ± standard deviation), indicating good performance.
• KGE components showed strong agreement in timing (average KGEr = 0.79 ± 0.08) and variability (average KGEg = 0.88 ± 0.22), with overall streamflow volume well-captured (average KGEb = 1.06 ± 0.30), though with some variability in bias.
• The method's robustness was confirmed across different lake and catchment sizes and configurations (Approach I.a, I.b, and II), with large lakes (Approach II) showing very good performance (e.g., KGE of 0.77 for Lake Vänern, 0.79 for Lake Vättern).
• Validation against observed data for 10 lakes yielded acceptable seasonal performance (KGE = 0.46 ± 0.21) and good annual performance (KGE = 0.70 ± 0.15), primarily limited by timing errors at the seasonal scale.

Contributions

• Establishes the first dynamic connection between the ISIMIP Global Water and Lake Sectors, addressing a critical limitation in global climate change impact assessments for lakes by integrating catchment-lake interactions.
• Develops and validates a robust and globally applicable scaling methodology for estimating streamflow inputs to individual lakes from coarse-resolution gridded hydrological model outputs.
• Provides a valuable dataset of estimated streamflow into lakes for the ISIMIP Lake Sector, which will support future hydrological and biogeochemical lake modeling efforts.
• Demonstrates the transferability and reliability of the scaling approach across a wide range of diverse hydrological and geomorphological settings.

Funding

• Vetenskapsrådet (grant no. 2021-04639)

Citation

@article{Ayala2026Integration,
  author = {Ayala, Ana I. and Hinostroza, José L. and Mercado-Bettin, Daniel and Marcé, Rafael and Gosling, Simon N. and Pierson, Donald C. and Sobek, Sebastian},
  title = {Integration of the Global Water and Lake Sectors within the ISIMIP framework through scaling of streamflow inputs to lakes},
  journal = {Geoscientific model development},
  year = {2026},
  doi = {10.5194/gmd-19-41-2026},
  url = {https://doi.org/10.5194/gmd-19-41-2026}
}