On the use of satellite, gauge, and reanalysis precipitation products for drought studies

Precipitation is a critical variable to monitor and predict meteorological drought. The WMO recommended standardized precipitation index (SPI) is calculated from gauge (i.e. GPCC), satellite-gauge (GPCP, CHIRPS), reanalysis (i.e. ERA-Interim, and MERRA-2), and satellite-gauge-reanalysis (i.e. MSWEP) over the global domain. Measured differences among the precipitation datasets include metrics such as percent area under drought, number of drought events, spread and correlation in the number of drought events, and critical success index in capturing moderate and severe-exceptional droughts. As precipitation products are available at different lengths and spatial resolutions, sensitivity of drought metrics to record-length and spatial resolution were explored. The results suggest that precipitation-based drought metrics can vary significantly with the choice of precipitation product, its record-lengths, and spatial resolution. These relationships also vary with the severity of drought events with more severe drought events being more sensitive to the differences in resolution and record length. The quantified variation among the products has to be recognized in the interpretation of drought events when a single or a subset of products used.