A probabilistic framework for sequential drought-fluvial identification, probability estimation and prediction

Abstract

• A probabilistic framework is proposed for evaluating sequential drought-fluvial event. • It can estimate the transition probability from prior droughts to succeeding fluvials. • It identifies changes in the joint dependence structure. • It predicts following fluvial severity based on the characteristics of prior droughts. While many modeling techniques have been developed to derive forecasts, frequency estimates and projection of droughts and floods, approaches that can inform how likely the two opposite extreme events could transit or propagate as cascades in time are still lacking. Here we proposed a probabilistic framework to serve the purpose. Specifically, the framework allows to 1) identify successive occurrence of hydrological droughts and fluvials, 2) estimate the probability of transition from a prior hydrological drought to a succeeding fluvial, and 3) make predictions of fluvials given a preceding dry spell. We applied this framework to predict the following fluvial severity based on the characteristics of hydrological drought at two runoff stations as an illustration. They are located in the Doingjiang Basin of Southern China (Station I) and the Eel River of United States (Station II), respectively. Our results show that the proposed framework can capture the observed fluvials based on the prior drought information at both stations, attesting to its skillful predictability. The transition probability estimate indicates that appreciable changes in the joint distribution of hydrological droughts and fluvials over time at Station I where the increased drought severity tends to enhance the occurrence likelihood of the fluvial happening in close succession, while Station II is subject to the opposite effect. Collectively, this study is expected to broaden the understanding of unusual shifts between droughts and floods, while providing useful information to help policy makers and infrastructure operators develop integrated approach to drought and flood management in the future.