Drought characteristics and prediction during pasture growing season in Xilingol grassland, northern China

In this paper, spatiotemporal variability of drought in Xilingol grassland during pasture growing season (from April to September) was investigated, using 52 years (1961–2012) of precipitation data recorded at 14 rain gauge stations in the study area. The Standardized Precipitation Index was used to compute the severity of drought. The Mann-Kendall test, the linear trend, and the sequential Mann-Kendall test were applied to standardized precipitation index (SPI) time series. The results indicate that drought has become increasingly serious on the region scale during pasture growing season, and the rate of SPI decreases ranged from −0.112 to −0.013 per decade. As for the MK test, most of the stations, the Z value range is from −1.081 to −0.005 and Kendall’s τ varies from −0.104 to −0.024. Meanwhile, drought is increased obviously from the northwest to the southeast region. Meanwhile, the occurrence probability of each severity class, times for reaching different drought class from any drought severity state, and residence times in each drought class have been obtained with Markov chain. Furthermore, the drought severities during pasture growing season in 2013–2016 are predicted depending on the weighted Markov chain. The results may provide a scientific basis for preventing and mitigating drought disaster.