Summary
This international model intercomparison study reveals that hydrological models with identical streamflow performance can exhibit substantial differences in internal process representation, particularly for interception, evaporation, snow dynamics and root-zone storage. By systematically comparing 12 calibrated models against multiple remote-sensing datasets, the authors demonstrate that streamflow alone is an insufficient evaluation metric and that model uncertainty extends well beyond observable discharge. The findings highlight fundamental trade-offs in model structure: models with small root-zone storage capacities risk overestimating dry-season stress, whilst those with large capacities tend to overestimate extreme dry conditions captured by satellite gravity anomalies.
UK applicability
The study's emphasis on multi-variable model validation is relevant to UK hydrological science and water resource management, particularly for the Meuse basin analogue in northwestern Europe. The findings suggest that UK-based hydrological models used for flood forecasting, drought assessment and water allocation should similarly be evaluated against diverse observational datasets beyond streamflow to improve confidence in internal process representation.
Key measures
Annual interception rates, seasonal evaporation rates, annual days with snow storage, mean annual maximum snow storage, root-zone storage capacity, remote-sensing estimates of evaporation (GLEAM), snow cover (MODIS), soil moisture, vegetation indices, and total storage anomalies (GRACE)
Outcomes reported
The study quantified differences in five states and fluxes across 12 process-based hydrological models that achieved similar streamflow performance, and assessed model plausibility using remote-sensing data on evaporation, snow cover, soil moisture and storage anomalies. Models showed substantial internal dissimilarities despite comparable streamflow outputs, with no single model consistently aligned across all available data sources.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.