Drought in the U.S. Great Plains (1980–2012): A sensitivity study using different methods for estimating potential evapotranspiration in the Palmer Drought Severity Index

Abstract

AbstractThe Palmer Drought Severity Index (PDSI) has been widely used to evaluate drought conditions since it was developed in 1965. In the original formulation of the PDSI, potential evapotranspiration (PET) was estimated using the Thornthwaite equation. This study evaluates how using more physically based approaches for estimating PET influences the depiction of drought conditions over the U.S. Great Plains from 1980 to 2012. Both the Penman‐Monteith equation and the two‐source PET model are compared to the original Thornthwaite‐based PDSI. The differences in PET between the three methods are much larger than the resulting differences in the PDSI. Results show that the original PDSI has a stronger drying trend than versions of PDSI that use more physically based methods of estimating PET. Spatially, all three versions of the PDSI show similar distributions of drying and wetting trends; however, there are significant regional variations that appear to be associated with land cover. PDSI and observed soil moisture in the top 1 m are moderately correlated (correlation coefficient is ~0.5) over the U.S. Great Plains, except in Texas (correlation coefficient is ~0.3). Although all three approaches result in a similar area‐averaged PDSI for the U.S. Great Plains, there are large differences in the area affected by drought, especially during extreme drought events.