Summary
This methodological study proposes spatial predictability of hydrological signatures as a criterion for their selection and ranking in catchment characterisation studies. Using three complementary approaches applied to the CAMELS dataset of 600+ U.S. catchments, the authors demonstrate that signatures exhibiting noisy spatial patterns tend to be poorly simulated, weakly linked to catchment attributes (particularly climate), and sensitive to data uncertainties. The findings provide guidance for practitioners selecting among the growing array of hydrological signatures for research and modelling applications.
UK applicability
The ranking framework and conclusions about signature uncertainty sensitivity are potentially transferable to UK catchments, though the CAMELS dataset is U.S.-based. Application to British hydrological conditions would require validation against UK catchment data and accounting for different climate regimes, soil types, and data collection protocols.
Key measures
Spatial predictability of hydrological signatures; random forest variable importance; Sacramento Soil Moisture Accounting model predictions; Moran's I spatial autocorrelation; sensitivity to discharge uncertainties
Outcomes reported
The study ranked 15 commonly used hydrological signatures based on their spatial predictability across 600+ U.S. catchments, using machine learning, hydrological simulation, and spatial autocorrelation analysis. Rankings revealed that signatures with noisy spatial patterns are poorly captured by hydrological models, weakly explained by climatic indices, and sensitive to discharge measurement uncertainties.
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.