A Generalized Density-Based Algorithm for the Spatiotemporal Tracking of Drought Events

Abstract

Abstract Drought events evolve simultaneously in space and time; hence, a proper characterization of an event requires the tracking of its full spatiotemporal evolution. Here we present a generalized algorithm for the tracking of drought events based on a three-dimensional application of the DBSCAN (density-based spatial clustering of applications with noise) clustering approach. The need for a generalized and flexible algorithm is dictated by the absence of a unanimous consensus on the definition of a drought event, which often depends on the target of the study. The proposed methodology introduces a set of six parameters that control both the spatial and the temporal connectivity between cells under drought conditions, also accounting for the local intensity of the drought itself. The capability of the algorithm to adapt to different drought definitions is tested successfully over a study case in Australia in the period 2017–20 using a set of standardized precipitation index (SPI) data derived from the ERA5 precipitation reanalysis. Insights on the possible range of variability of the model parameters, as well as on their effects on the delineation of drought events, are provided for the case of meteorological droughts in order to incentivize further applications of the methodology.