Understanding drought indicator-to-impact relationships to improve drought monitoring and early warning: Thailand as a case study

Abstract

Thailand has already been experiencing an increase in severity and duration of its droughts as a consequence of the changing climate. Developing a reliable drought monitoring and early warning system (DMEWS) is an integral part of strengthening a country’s resilience to droughts. However, for DMEWS to be useful for stakeholders, the indicators they monitor should be translatable to potential drought impacts on the ground and, ideally, inform mitigating actions. Here, we analyse these drought indicator-to-impact relationships in Thailand, using a novel combination of correlation analysis and random forest modelling. In the correlation analysis, we study the link between meteorological drought indicators and high-resolution remote sensing vegetation indices used as proxies for general crop health and forest growth. Our analysis shows that these links vary greatly depending on land use (cropland vs. forest), season (wet vs. dry) and region (north vs. south). The random forest models built to estimate regional crop productivity provided a more in-depth analysis of the crop- and region-specific value of different drought indicators. The results highlighted seasonal patterns of drought vulnerability for individual crops, usually linked to their growing season, although the effect was somewhat masked in irrigated regions (North). This new high-resolution knowledge of crop- and region-specific indicator-to-impact links can be used as the basis of targeted mitigation actions in an improved DMEWS in Thailand. In addition, the framework developed here can be applied elsewhere in the Southeast Asia region, as well as other drought-vulnerable areas internationally, in particular those that are data sparse.