Seasonal reforecast skill assessment in Sweden

Résumé fourni par la source

This dataset contains results from an assessment of the seasonal predictability of streamflow extremes across Sweden. The analysis is based on hydrological reforecasts produced with the S-HYPE hydrological model, driven by the ECMWF SEAS5 seasonal climate reforecast ensemble, with S-HYPE driven by PTHBV as reference. The reforecast period covers 1993-2015 and includes more than 39,000 catchments across Sweden. Forecast skill is evaluated for lead times from lead week 0 to lead week 29. The analysis focuses specifically on the predictability of low- and high-streamflow extremes. Extreme-flow events are defined using percentile-based thresholds derived from the streamflow distribution: 5th percentile (Q05): very low streamflow 10th percentile (Q10): low streamflow 90th percentile (Q90): high streamflow 95th percentile (Q95): very high streamflow Forecast skill is calculated separately for each catchment, lead week, and extreme-flow threshold. The skill assessment follows the methodological framework described in Popp et al and saved separately as bsXX.skill.csv where XX denotes each extreme threshold. The CSV files contain the calculated forecast-skill using Brier Skill Score at each: S-HYPE catchment (SUBID) , row forecast lead week (0–29 weeks), column

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.