Use of four drought indices for evaluating drought characteristics under climate change in Shaanxi, China: 1951-2012

Drought severity was simulated with four drought indices to examine the impacts of climate change on drought conditions in Shaanxi province over the period 1951 to 2012. The drought metrics analyzed were based on the original Palmer drought severity index (orPDSI), self-calibrated PDSI (scPDSI), the standardized precipitation index (SPI) and the standardized precipitation evapotranspiration index (SPEI). Both Thornthwaite (Thor) and Penman-Monteith (PM) parameterizations were used to calculate potential evapotranspiration (PET), and the differences between two PET estimators were studied. Nonparametric Mann-Kendall monotonic test was used to examine the trends of hydroclimatic data. Series of drought indices were compared at five meteorological stations with different climate characteristics, located in the north, central and south parts of Shaanxi province, respectively. Effects of climate change in drought conditions were investigated with hypothetical progressive precipitation decrease (-15 %) and temperature increase (2 A degrees C). The results showed that there was discrepancy between PET estimated using the Thor and PM parameterization estimators, while the SPEI calculated with the two PET estimators are found to be similar. The SPEI has the combined advantages over the scPDSI and the SPI, considering the effect of temperature variability on drought severity and its multi-scalar characteristic, while scPDSI has an inherent approximately 12-month time scale. The Pearson's correlation is used to compare the three pairs of drought indices combinations at different time scales. Under climate change conditions, the drought severity increases with the decline of precipitation and higher water demand as a result of the temperature increase based on the metrics of the scPDSI, the SPI and the SPEI.