Methodology:

This data can be produced using a high-resolution (100 m, daily timestep) physically based hydrology model comprising four modules: supraglacial routing and lake-filling, hydrofracture, slow lake drainage (overtopping), and lake refreezing. Meltwater routing is simulated using a DEM-based flow-path algorithm; hydrofracture using linear elastic fracture mechanics; slow drainage via channel incision parameterization; and refreezing with an energy-balance approach. We provide the data to run the model over three distinct catchments in southwest Greenland (regions 1-3) from 2015 to 2019. The model parameters (e.g., critical stress threshold, channel width) have been calibrated by comparing model output to satellite-observed lake areas and in situ discharge.

Data collection:

Key inputs include:

- ArcticDEM (100 m surface elevation) - for hydrological routing and lake formation.

- MAR v3.11 regional climate model output - daily surface runoff, snow depth/density, meteorological forcing (air temperature, humidity, wind, shortwave radiation) at approximately 6 km resolution.

- Surface von Mises stress fields - for crevasse/hydrofracture initiation.

- SHED v1.0 model code (Fortran-77).

- Ancillary scripts and setup files for running SHED modules and post-processing.

Data quality:

The SHED v1.0 model has been evaluated against multiple independent observational datasets. Modelled supraglacial lake locations and temporal evolution were compared with lake extents derived from MODIS satellite imagery, showing that approximately 80 % of observed lake locations were reproduced and that seasonal lake area dynamics were captured with reasonable accuracy. In addition, simulated meltwater discharge routed from the supraglacial system was compared with in situ proglacial discharge measurements from southwest Greenland catchments, demonstrating good agreement in magnitude and seasonal timing, subject to expected delays due to the absence of an explicit subglacial hydrology component. Key model parameters were calibrated using one study region and subsequently applied to other regions to assess transferability. Uncertainties related to input data (e.g. DEM accuracy, climate model runoff) and simplified process representations remain, but overall validation results indicate the model is robust for simulating supraglacial hydrology and drainage at regional scales.