Pulse Brain · Growing Health Evidence Index
Tier 3 — Observational / field trialPeer-reviewed

A Ranking of Hydrological Signatures Based on Their Predictability in Space

Nans Addor, Grey Nearing, Cristina Prieto, Andrew J. Newman, Nataliya Le Vine, Martyn Clark

Water Resources Research · 2018

Read source ↗ All evidence

Summary

This paper addresses the lack of guidance for selecting hydrological signatures by proposing that spatial predictability provides a robust criterion for evaluation. Using three complementary approaches across the CAMELS dataset of 600+ U.S. catchments, the authors ranked 15 commonly used hydrological signatures and demonstrated that signatures with poor spatial structure are poorly simulated, weakly related to catchment attributes, and sensitive to measurement noise—findings with implications for catchment classification, process understanding, and model calibration studies.

Regional applicability

The methodological framework and ranking approach could be applied to UK catchments and hydrological monitoring networks to improve signature selection for national and regional water resource management and flood characterisation studies. However, the specific signature rankings would need validation against UK catchment attributes and hydroclimate conditions.

Key measures

Random forest feature importance scores; spatial autocorrelation (Moran's I); Sacramento Soil Moisture Accounting model predictions; 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 modelling, and spatial autocorrelation analysis. Signatures with noisier spatial patterns were found to be poorly captured by hydrological simulations, less explained by catchment attributes, and more sensitive to discharge uncertainties.

Theme
Measurement & metrics
Subject
Measurement methods & nutrient profiling
Study type
Research
Study design
Comparative analysis using machine learning and hydrological modelling
Source type
Peer-reviewed study
Status
Published
Geography
United States
System type
Other
DOI
10.1029/2018wr022606
Catalogue ID
BFmommpl0r-4r81n6

Topic tags

Pulse AI · ask about this record

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.