A drought monitoring framework for data-scarce regions

Drought monitoring is a critical activity for drought risk management; however, the lack of ground-based observations of climatological and hydrological variables in many regions of the world hinders an adequate follow-up and investigation of this phenomenon. This paper introduces a transparent framework for monitoring the spatio-temporal distribution of drought hazard based on uni- and multivariate standardized drought indices that use reanalysis datasets of hydrological variables available freely and globally. In the case study of the 2015–2017 East-Southwest drought in Mexico, the introduced framework successfully detected the spatial and temporal patterns of drought conditions, even in regions where a benchmark drought monitoring system failed to detect deficits. In addition, the ability of the introduced framework to detect drought impacts on the annual agricultural maize production in Mexico was evaluated using data of 1980–2018, yielding scores of the false alarm ratio =0.32, the probability of detection = 0.71, and the proportion correct = 0.68 for the analysis at the national scale. Currently, the framework provides a significant extension to the capabilities for national drought monitoring, and it is being used by the Mexican water authority in the decision-making process related to drought severity assessment.