Projections of water requirements of cotton and sugar beet in Xinjiang based on statistical downscaling model

Abstract: Information on the spatial distribution characteristics of crop water requirements in Xinjiang Weiwuer Autonomous Region under the climate change scenarios can be used as references for the agricultural water use planning. Previous research on climate change mainly focuses on the temporal changes analysis of temperature, precipitation, or reference crop evapotranspiration under certain emission scenarios. The effects of climate change on variations of crop water requirements in Xinjiang are limited because the methods are more complex and require participation of multi-disciplinary team. In this study, daily meteorological data from 41 weather stations were collected from 1961 to 2010 in Xinjiang region. There were typical geographical and meteorological differences among the selected 41 sites. Reference crop evapotranspiration and crop coefficients were calculated for each site using the FAO-56 Penman-Monteith equation and the single-crop coefficient method, respectively. Crop water requirement of cotton and sugar beet were obtained through reference crop evapotranspiration multiplied by crop coefficient. The statistical downscaling model (SDSM 4.2 software) was applied to each site of the region to project daily crop water requirement sequence of cotton and sugar beet from year 2015 to 2099 under two emission scenarios of high and low emission. The results showed that crop water requirement for cotton and sugar beet ranged from 0.11 to 1.04 at the initial growth stage, 0.98 to 1.05 at the middle growth stage, and 0.58 and 0.7 at the final growth stage during the period of 1961 to 2010 in Xinjiang. The multi-year average reference crop evapotranspiration over 1961-2010 in Xinjiang area varied from 1.84-2.76 mm/d, and the difference in reference crop evapotranspiration at different geographical locations was obvious. The spatial distributions of multi-year average crop water requirement for cotton and sugar beet during growth stage decreased gradually from the south to the north. crop water requirement of cotton varied from 573.9 to 2 853.2 mm/a and generally decreased from the south to the north of Xinjiang except stations of Yiwu and Wuqia in which there were large values of crop water requirement. Crop water requirement of sugar beet varied from 261.4 to 1 300.8 mm/a and the spatial distributions crop water requirement of sugar beet also decreased from the south to the north gradually. For SDSM projections, among the 26 predictors, the average surface specific humidity and average surface temperature respectively correlated to crop water requirement well (P<0.01) at most of the sites. During 2015-2099 under high and low emission scenarios, multi-year mean crop water requirement values of sugar beet in Xinjiang area were from 93.4 to 576.9 and 87.4 to 574.5 mm/a. The projected multi-year mean crop water requirement values of cotton were from 303.6 to 1 608.0 and 305.0 to 1 640.4 mm/a, which were all smaller than those calculated for the period of 1961-2010. Crop water requirement under both emission scenarios decreased gradually from the south to the north. The spatial distributions of crop water requirement from 2015 to 2099 under high and low emission scenarios were similar to that of crop water requirement over 1961-2010, but were generally smaller in values. Overall, crop water requirement of cotton and sugar beet during the historical and future periods decreased at various extents. This research provides possible changes of crop water requirement in the future in Xinjiang region, which can be useful as a reference for irrigation decision and agricultural water management planning.