Global characterisation of land-atmosphere interactions during flash droughts using satellite observations

Land-atmosphere interactions are known to be important for the development of flash droughts, and improving the representation of these interactions in subseasonal-to-seasonal (S2S) forecasting models would provide a potential source of skill for predicting these events. However, understanding the land-atmosphere coupling processes involved in flash drought development globally is hindered by the fact that key variables such as root-zone soil moisture and surface latent and sensible heat fluxes cannot be directly observed from satellites. In this study, we use a definition of flash droughts based on ESA CCI soil moisture to explore the composite behaviour of land-atmosphere variables around flash drought onset dates. We exploit satellite-observed land surface temperature (LST) data from ESA CCI to diagnose the balance between latent and sensible surface heat fluxes by computing the difference between LST and 2m air temperature (T2m) from ERA5 reanalysis. Since the standardised anomaly of the sensible heat flux is approximately equal to the standardised anomaly of LST-T2m, this method allows us to identify increases in sensible heat flux anomalies during flash droughts. When radiation conditions remain approximately constant, this is associated with the onset of a water-limited evaporative regime. We explore the spatial variation in the sensitivity of both LST-T2m and Vegetation Optical Depth (VOD) to flash drought events, to understand where the surface energy budget changes most strongly and where impacts on vegetation are most severe. Additionally, we consider which satellite-observable variables are most promising for providing information that can improve the S2S prediction of flash droughts.