Towards robust seasonal streamflow forecasts in mountainous catchments: impact of calibration metric selection in hydrological modeling

Abstract. Dynamical (i.e., model-based) methods are widely used by forecasting centers to generate seasonal streamflow forecasts, building upon process-based hydrological models that require parameter specification (i.e., calibration). Here, we investigate the extent to which the choice of calibration objective function affects the quality of seasonal (spring-summer) streamflow forecasts produced with the traditional ensemble streamflow prediction (ESP) method and explore connections between forecast skill and hydrological consistency – measured in terms of biases in hydrological signatures – obtained from the model parameter sets. To this end, we calibrate three popular conceptual rainfall-runoff models (GR4J, TUW, and Sacramento) using 12 different objective functions, including seasonal metrics that emphasize errors during the snowmelt period, and produce hindcasts for five initialization times over a 33-year period (April/1987–March/2020) in 22 mountain catchments that span diverse hydroclimatic conditions along the semiarid Andes Cordillera (28°–37° S). The results show that the choice of calibration metric becomes relevant as the winter (snow accumulation) season begins (i.e., July 1), enhancing inter-basin differences in forecast skill as initializations approach the beginning of the snowmelt season (i.e., September 1). The comparison of seasonal forecasts obtained from different calibration metrics shows that hydrological consistency does not ensure satisfactory seasonal ESP forecasts (e.g., Split KGE), and that satisfactory ESP forecasts are not necessarily associated to a hydrologically consistent parameter set (e.g., VE-Sep). Among the options explored here, an objective function that combines the Kling-Gupta Efficiency (KGE) and the Nash-Sutcliffe Efficiency (NSE) with flows in log space provides the best compromise between hydrologically consistent model simulations and good forecast performance. Finally, the choice of calibration metric generally affects the magnitude of correlations between forecast quality attributes and catchment descriptors, rather than the sign, being the baseflow index and interannual runoff variability the best predictors of forecast skill. Overall, this study highlights the need for careful parameter estimation strategies in the forecasting production chain to generate skillful forecasts for the right reasons and draw robust conclusions on hydrologic predictability.