Chan et al. (2026) UK Hydrological Outlook using Historic Weather Analogues

Identification

• Journal: Hydrology and earth system sciences
• Year: 2026
• Date: 2026-02-17
• Authors: Wilson Chan, Katie A. Facer-Childs, Maliko Tanguy, Eugene Magee, Burak Bulut, Nicky Stringer, Jeff Knight, Jamie Hannaford
• DOI: 10.5194/hess-30-905-2026

Research Groups

• UK Centre for Ecology & Hydrology (UKCEH), Wallingford, UK
• European Centre for Medium-Range Weather Forecasts (ECMWF), Reading, UK
• Met Office, Exeter, UK
• Irish Climate Analysis and Research UnitS (ICARUS), Maynooth University, Maynooth, UK

Short Summary

This study assesses the skill of a new Historic Weather Analogues (HWA) method for seasonal hydrological forecasts across 314 UK catchments, benchmarking it against the standard Ensemble Streamflow Prediction (ESP) and climatology. The HWA method significantly improves winter river flow forecasts nationally, especially in upland, fast-responding catchments, while maintaining comparable skill to ESP in other regions and seasons.

Objective

• When is the HWA forecasts skilful, across different seasons?
• Where are the HWA forecasts skilful, across UK catchments and regions?
• To what extent are the HWA forecasts an improvement against the standard ESP approach?

Study Configuration

• Spatial Scale: 314 UK catchments, representing a wide range of hydroclimatic conditions across the UK (national scale).
• Temporal Scale: Seasonal (3-month lead time) hydrological hindcasts over the 1993–2016 period for four meteorological seasons (DJF, MAM, JJA, SON).

Methodology and Data

• Models used:
  • Hydrological model: GR6J (Génie Rural à 6 paramètres Journalier) version 1.0.2
  • Weather forecasting model: GloSea6 (Met Office Hadley Centre climate model HadGEM3)
• Data sources:
  • Observed daily river flow: National River Flow Archive (NRFA)
  • Observed rainfall and temperature: 1 km HadUK-Grid dataset
  • Potential evapotranspiration (PET): Calculated using daily average temperature with the McGuinness-Bordne equation.
  • Atmospheric circulation indices (NAO): University of East Anglia’s Climatic Research Unit (CRU)
  • North Atlantic jet stream intensity: ERA5 reanalyses

Main Results

• The HWA method significantly improves river flow forecasts in winter, enabling skilful winter river flow forecasts across the UK, whereas the standard ESP method was only skilful in southeast England.
• Winter river flow forecasts using the HWA method were more skilful in discriminating high and low flows across all regions compared to ESP.
• Catchments showing the greatest improvement with HWA tended to be upland, fast-responding catchments with limited catchment storage, where river flow variability is strongly tied to climate variability.
• Skilful winter river flow predictability with HWA is attributed to the relatively high forecast skill of winter atmospheric circulation patterns (e.g., NAO) and HWA's ability to derive high-resolution meteorological inputs.
• Improvement in river flow forecast skill for other seasons was modest; moderate improvements were observed in northern England and northeast Scotland during spring, with little change in autumn.
• Skilful summer flow predictability remains possible only for southeast England, and skill scores at some catchments were reduced compared to the ESP method in summer.
• HWA skill improvement in winter is negatively correlated with the Baseflow Index (BFI) (r = -0.44) and positively correlated with Standardised Annual Average Rainfall (SAAR) (r = 0.51) and the catchment wetness index (r = 0.64).
• Nationally, 45% of catchments showed a positive Continuous Ranked Probability Skill Score (CRPSS) and 64% showed a positive Ranked Probability Skill Score (RPSS) for HWA compared to ESP in winter.

Contributions

• Provides the first comprehensive, catchment-based assessment of the Historic Weather Analogues (HWA) method for seasonal hydrological forecasting across the UK, benchmarking its skill against the widely used Ensemble Streamflow Prediction (ESP) and climatology.
• Demonstrates that the HWA method significantly enhances winter river flow predictability nationally, particularly for fast-responding catchments in northern and western UK, by effectively leveraging climate information from dynamical weather forecasting models and exploring historically unseen weather sequences.
• Identifies the specific spatial and seasonal contexts where the HWA method offers a clear improvement over existing forecasting approaches, while also confirming the standard ESP method's robustness as a benchmark.

Funding

• UKCEH National Capability for UK Challenges Programme [NE/Y006208/1]
• HydroJULES programme [NE/S017380/1]
• Natural Environment Research Council (NERC) Climate change in the Arctic–North Atlantic Region and Impacts on the UK (CANARI) project [NE/W004984/1]

Citation

@article{Chan2026UK,
  author = {Chan, Wilson and Facer-Childs, Katie A. and Tanguy, Maliko and Magee, Eugene and Bulut, Burak and Stringer, Nicky and Knight, Jeff and Hannaford, Jamie},
  title = {UK Hydrological Outlook using Historic Weather Analogues},
  journal = {Hydrology and earth system sciences},
  year = {2026},
  doi = {10.5194/hess-30-905-2026},
  url = {https://doi.org/10.5194/hess-30-905-2026}
}