Drought indicators and triggers: A stochastic approach to evaluation

Drought management depends on indicators to detect drought conditions, and triggers to activate drought responses. But determining those indicators and triggers presents challenges. Indicators often lack spatial and temporal transferability, comparability among scales, and relevance to critical drought impacts. Triggers often lack statistical integrity, consistency among drought categories, and correspondence with desired management goals. This article presents an approach for developing and evaluating drought indicators and triggers, using a probabilistic framework that offers comparability, consistency, and applicability. From that, a multistate Markov model investigates the stochastic behavior of indicators and triggers, including transitioning, duration, and frequency within drought categories. This model is applied to the analysis of drought in the Apalachicola-Chattahoochee-Flint River Basin in the southeastern United States, using indicators of the Standardized Precipitation Index (for 3, 6, 9, and 12 months), the Palmer Drought Severity Index, and the Palmer Hydrologic Drought Index. The analysis revealed differences among the performance of indicators and their trigger thresholds, which can influence drought responses. Results contribute to improved understanding of drought phenomena, statistical methods for indicators and triggers, and insights for drought management.