A Comparison of the Standardized Precipitation Evapotranspiration Index Using Thronthwaite and Penman-Monteith Parameterizations for Potential Evapotranspiration

The Standardized Precipitation Evapotranspiration Index (SPEI) is a drought index which incorporates a simple climatic water balance and requires precipitation and potential evapotranspiration (PET) as inputs. The simple Thornthwaite approach is commonly utilized in the SPEI to calculate PET because the only requirement is mean monthly temperature. To potentially improve the ability of SPEI identifying drought conditions a more physically based PET approach such as the Penman-Monteith is desired, however, temperature, wind speed, vapor pressure, and solar radiation are required. This study uses a recently developed high resolution (4-km) bilinearly interpolated NLDAS-2 gridded dataset to compare two versions of SPEI using the Thornthwaite (SPEI-Th) and Penman Monteith (SPEI-PM) approaches calculated at a variety of time scales over the contiguous United States (CONUS) from 1979 through 2010. SPEI-Th and SPEI-PM were both correlated to monthly and water-year standardized streamflow from three different mountainous regions of the country with contrasting climates. The three regions include southwestern California (hot and dry climate), the Cascade Range in Washington (cool and moist climate), and the Great Smokey Mountains of Tennessee/North Carolina (humid continental climate). As expected, major differences in PET rates are found between the two PET models, however, SPEI-Th and SPEI-PM are very similar over much of the CONUS at all time scales. The largest differences between SPEI-Th and SPEI-PM are found over the desert Southwest at short time scales. While using the physically based Penman-Monteith approach leads to slightly higher correlations to streamflow, especially in water limited regions, our results suggest that the Throntwaite approach provides similar SPEI sensitivity in energy limited regions. These results indicate that it may be more useful to use a physically based PET approach for drought monitoring, especially in water limited regions.