How well do hydrological models simulate streamflow extremes and drought-to-flood transitions?

Abstract. The impacts of floods can be enhanced when they occur shortly after drought. Models can be a useful tool to better understand the processes and mechanisms driving the response of floods occurring in close succession to streamflow drought. However, it is yet unclear how well hydrologic models capture these compound extreme events and which modeling decisions are most important for high model performance. To address this research gap, we calibrated four conceptual bucket-type hydrological models with different structures (GR4J, GR5J, GR6J, and TUW) for 63 catchments in Chile and Switzerland using different calibration strategies. We tested different configurations of the Kling-Gupta efficiency (KGE) formulation for model calibration to assess their performance in simulating and detecting observed transitions. We assessed the relative importance of different methodological choices including model structure, streamflow transformation, and KGE formulation and weights. We demonstrate that model performance as expressed by the KGE or NSE does not guarantee a good performance in terms of detecting streamflow extremes and their transitions. Furthermore, we show that a model's performance with respect to capturing extreme events primarily depends on how well it captures streamflow timing (i.e., correlation between observations and simulations) rather than other hydrological signatures or variables such as evapotranspiration or snow water equivalent. Our results also highlight that model structure and catchment characteristics as well as meteorological forcing play a key role in the detection of transitions. Specifically, we demonstrate that drought-to-flood transitions are more difficult to capture in semi-arid high-mountain catchments than in humid low-elevation catchments. Finally, our study provides guidelines for further model improvements with respect to drought-to-flood transitions, which can support process understanding related to these compound events, identifying regions prone to this type of event, and contribute to improved risk management -- aspects that will enhance preparedness.