Evaluating the applicability of exclusively optical sensors in flash drought monitoring with MODIS observations

Remote sensing of flash drought (FD) detection with fine spatial resolutions provides better understanding and applications to regional managements related to meteorology, hydrology, and agriculture. The surface evaporative stress via thermal observations from space is usually used for detecting FD, however, thermal bands are not provided in every advanced satellite and usually face limitations over rugged landscape at a fine resolution. This study therefore analyzes the applicability of optical-only sensors in monitoring FD by calculating standardized evaporative stress ratio (optical-based SESR) without the need of thermal wavelengths, based on the trapezoidal concept. A total of 14-year dataset from the MOD09A1 products was acquired for the analysis over the southeast Minnesota and central Wisconsin, United States. Results showed that optical-based SESR was positively correlated (p < 0.01) with Vegetation Health Index (VHI) and Temperature Condition Index spatiotemporally across the study regions. Monthly mean values of optical-based SESR successfully captured spatiotemporal drought evolution over the two study regions during the growing season in 2012. The optical-based SESR possibly provides early warning of FD when vegetation starts exhibiting worsening signals (VHI < 40 %) and precipitation are unusually low (precipitation anomalies < –1.0). In both study areas, intensification of FD in 2012 and its local characteristics driven by land–atmosphere coupling were fully reflected by optical-based SESR. Climatological analysis of optical-based SESR found that high frequency of FDs are likely to take place over agricultural land and tend to occur more frequently in recent years. This study provides a comprehensive evaluation for optical sensors and suggests a suitable alternative for thermal-/microwave-based FD detections to regional-scale studies, with a versatile applicability to various satellite observations.