Bias-corrected and spatially disaggregated seasonal forecasts: a long-term reference forecast product for the water sector in semi-arid regions

Abstract. Seasonal forecasts have the potential to substantially improve water management particularly in water-scarce regions. However, global seasonal are usually not directly applicable as they provided at coarse spatial resolutions of best 36 km and suffer from model biases drifts. In this study, we therefore apply a bias-correction spatial-disaggregation (BCSD) approach precipitation, temperature radiation latest long-range forecasting system SEAS5 European Centre for Medium-Range Weather Forecasts (ECMWF). As reference use data ERA5-Land offline land surface rerun ECMWF reanalysis ERA5. Thereby, correct drifts resolution 0.1?. This is performed example over four predominately semi-arid study domains across world, which include river basins Karun (Iran), São Francisco River (Brazil), Tekeze–Atbara Blue Nile (Sudan, Ethiopia Eritrea), Catamayo–Chira (Ecuador Peru). Compared against ERA5-Land, bias-corrected spatially disaggregated higher show reduced better agreement patterns than raw well remarkably lead-dependent drift effects. But our analysis also shows that computing monthly averages daily during periods with strong temporal climate gradients or heteroscedasticity can lead remaining especially lowest- highest-lead forecasts. Our BCSD cover whole (re-)forecast period 1981 2019 ensemble average, minimum, maximum shortwave issue date next 215 d 6 months, respectively. sums up more 100 000 forecasted days each 25 (until year 2016) 51 (from 2017) members five analyzed variables. The full repository made freely available public via World Data Climate https://doi.org/10.26050/WDCC/SaWaM_D01_SEAS5_BCSD (Domain D01, Basin Lorenz et al., 2020b), https://doi.org/10.26050/WDCC/SaWaM_D02_SEAS5_BCSD D02: 2020c), https://doi.org/10.26050/WDCC/SaWaM_D03_SEAS5_BCSD D03: (Ethiopia, Eritrea, Sudan), 2020d), https://doi.org/10.26050/WDCC/SaWaM_D04_SEAS5_BCSD D04: (Ecuador, Peru), 2020a). It currently first publicly high-resolution forecast product covers multiple regions variables such long period. hence provides unique test bed evaluating performance driving hydrological, ecosystem impact models. Therefore, provide crucial contribution disaster preparedness and, finally, proofing regional climatically sensitive