Linear and Non-Linear Characteristics and Determinants of Flash Droughts from MERRA2 and ERA5-Land

Although, there is some understanding of what a flash drought (FD) might "look" like (e.g., precipitation deficits, and high temperatures), there is still some ambiguity on what constitutes a FD, i.e., either rapid intensification or duration or both. A precipitation deficit is a basic requirement for drought to develop, the speed with which it develops, and its ultimate severity are influenced by other environmental anomalies (temperatures, evapotranspiration, humidity, winds, radiations etc.,). However, the incongruities that exist within literature on FD definitions, methodologies, datasets, and data sources used will influence the FD characteristics. In this work, we synthesize the research to-date on FD definitions and methodologies using two major data sources MERRA2 (1980-2021) and ERA5-Land (1950-2021) to characterize the uncertainties in FD characteristics. In addition, a global sensitivity analysis is conducted to identify the linear (first order) and non-linear (second order) impacts of the environmental drivers on FD characteristics both spatial and temporal (areal extent, count, development time, duration). The period used to identify a FD will be determined using climatological NDVI. Primarily, two main global FD features are quantified: mean time scale for FD development and sustenance, and the rate of change in FD (growth, decline, expansion, and shrinkage).