Vapor pressure deficit (VPD) is an important parameter for featuring atmospheric drought, which plays a significant role in analyzing the impact of spatio−temporal VPD variations on the climate change and agricultural and animal husbandry production. The relevant monthly meteorological data, including mean air temperature(Tm), mean maximum air temperature, mean minimum air temperature, sunshine hours, precipitation, relative humidity, vapor pressure and mean wind speed, were collected at 38 meteorological stations in Xizang from 1981 to 2020. Several other relevant corresponding variables, such as the indices of grain yield, atmospheric circulation and sea surface temperature (SST), were also collected accordingly. Based on these materials, both the spatio−temporal characteristics of VPD variation and the influential factors were analyzed to identify their impacts on grain yield over Xizang in the recent 40 years, with the support of different statistical methods such as linear tendency estimation, Pearson's correlation coefficient, Pettitt mutation test method, linear tendency estimation and stepwise regression etc. The results indicated that: (1) spatially, there were 3 regions with high annual mean VPD in Xizang, which were located in the western part of Ngari district, the agricultural area along the Yarlung Zangbo river valley, and Baxoi station in the dry heat valley of the Nujiang river, respectively; the regions with low values were distributed in the east−central part of the Nagchu and at the southern edge of Xizang. In the recent 40 years, the annual VPD at all stations increased at varying rates of 0.014−0.064kPa·10y−1 (13 stations P<0.01), with the largest value in Lhasa, followed by Xigazê (0.044kPa·10y−1, P<0.001), and the smallest in Lhari. Overall, the linear trend of VPD decreased with the increase in altitude. (2) Temporally, the annual VPD in Xizang fluctuated greatly, showing a 'V' type change in the past 40 years. Initially, the VPD exhibited a significant decreasing trend from 1981 to 2000 at a rate of −0.020kPa·10y−1 (P<0.05), then the VPD increased significantly with a rate of 0.053kPa·10y−1 (P<0.01) from 2001 to 2020. Overall, the VDP showed an increasing trend of 0.025kPa·10y−1 during 1981−2020. In addition, a sudden increase in both annual and seasonal VPD occurred around 2005. Both the annual and the seasonal VPD were lower in 1980s and 1990s, especially in the 1990s. In 2000s, the VPD showed lower values in spring and autumn, higher values in summer and winter, with slightly higher annual values. In contrast, both annual and seasonal values of VPD were higher in 2010s, especially in the summer and autumn times. (3) Except for the summer time, both seasonal and annual average VPD were positively correlated with grain yield significantly, especially with the yield per unit area in Xizang, which reached a very significant test level of P<0.001. The VPD in growing season was only positively correlated with total grain yield significantly, and VPD had the most significant effect on total yield and yield per unit area in April and September. The increase of VPD was beneficial to grain yield enhancement. In contrast, for maize and soybean cultivated in northeastern and northern China, crop yields decreased along with increasing VPD. Therefore, unlike other agricultural production regions in China, climate change might play a positive role in crop growth and yield formation through increasing VPD, meaning that water−saving technologies and agronomic management need to be strongly encouraged to adapt to the ongoing climate change in Xizang. (4) Different dominant factors were analyzed for meteorological, circulation and SST impacts on VDP variations over Xizang in the recent 40 years. The increase in seasonal and annual Tm was identified as the dominant factor for the meteorological impact, while the Xizang plateau region 1 index was determined to be the dominant circulation factor impacting annual and seasonal VPD variations. As for the dominant factors for SST impact, the Indian ocean warm pool area index was believed to be the dominant SST factor affecting VPD variations in spring and winter, while the Indian ocean basin−wide index was found to be the dominant SST factor affecting VPD variations in summer, autumn and all year around. All of the dominant factors mentioned above have shown a significant increasing trend in the last 40 years, resulting in a significant increase of VPD in Xizang accordingly.
