Uncertainties in drought from index and data selection

This study compares various drought indices, the Palmer Drought Severity Index (PDSI), the Standardized Precipitation Evapotranspiration Index (SPEI), and the Standardized Precipitation Index (SPI), and the uncertainties in each to the input data from which they are derived. The abilities of the PDSI, SPEI, and SPI to capture drought periods are assessed through a comparison with soil moisture estimates from two generations of the Global Land Data Assimilation System (GLDAS). This comparison shows that the skill with which a drought index represents variations in soil moisture does not necessarily improve when evapotranspiration is included (i.e., PDSI/SPEI rather than SPI), though this depends on location and the time scale of the drought. The differences in the abilities of the drought indices to represent soil moisture are also compared to the magnitude of the uncertainty in each index arising from the choice of input data. In many cases, the uncertainties in the variations of the PDSI, SPEI, and SPI to the choice of input data are larger in magnitude than the differences between the indices themselves, particularly when considering the dry tails of the distribution. The results show that no one drought index outperforms the others during drought conditions.