Comparison of Deterministic and Probabilistic Variational Data Assimilation Methods Using Snow and Streamflow Data Coupled in HBV Model for Upper Euphrates Basin

The operation of upstream reservoirs in mountainous regions fed by snowmelt is highly challenging. This partly due to scarce information given harsh topographic conditions and a lack monitoring stations. In this sense, snow observations from remote sensing provide additional relevant about the current basin. can be used improve model states forecast using data assimilation techniques, therefore enhancing reservoirs. Typical techniques effectively reduce uncertainty initialization merging simulations observations. However, they do not take into account model, structural, or parametric uncertainty. intrinsic introduces complexity restricts daily work operators. novel Multi-Parametric Variational Data Assimilation (MP-VarDA) uses different parameter sets create pool models that quantify arising parametrization. study focuses on sensitivity reduction MP-VarDA coupled HBV hydrological pools impact number performance streamflow Snow Cover Area (SCA) forecasts. created Monte Carlo simulation, combined with an Aggregated Distance (AD) Method, instances. tests are conducted Karasu Basin, located at uppermost part Euphrates River Türkiye, where significant portion yearly runoff. analyses were for thresholds based observation exceedance probabilities. According results comparison deterministic VarDA, probabilistic improves m-CRPS gains forecasts 57% 67% BSS skill 52% 68% when SCA assimilated. improvement rapidly increases first but reaches maximum benefit after 5 pool. notable both methods forecasts, best gain obtained VarDA (31%), while detected (12%).