Evaluation and projection of a snow coverage rate over the Upper Yarkant River Basin, China

Based on snow coverage remote sensing data NOAA, the simulation ability of the snow coverage rate from the core experiments of the 5th Coupled Model Intercomparison Project (CMIP5) over the Upper Yarkant River Basin was evaluated. The ensemble mean of the eight models was used to project future changes in the snow coverage rate over the basin under the three different representative concentration pathways (RCP2.6, RCP4.5 and RCP8.5). The results show that the snow coverage rate of the eight models can basically simulate a sharp drop of the spatial distribution from the west to the northeast; all the eight models have the capacity to simulate the snow coverage rate of the Yarkant River Basin analyzed from the correlation analysis (R), the ratio of spatial standard deviation (SDR) and the S index. The simulation capacity of the eight models in spring, autumn and winter is stronger than the one in summer, and the ensemble mean of the eight models simulates the best. Under the scenarios of RCP2.6 and RCP4.5, changes in the snow coverage rate show negative differences in most areas over the basin, only the local part shows positive differences. Under the scenario of RCP8.5, changes in the snow coverage rate in all seasons present negative differences in the whole basin, and the decreasing trends are more obvious. Under the same emission scenario, the negative differences in winter are weaker than those of other seasons, and the most significant negative difference appears in spring and autumn. The projection scenarios indicate a clearly decreasing trend under the different emission scenarios. Under the scenario of RCP2.6, the decreasing trend is not obvious as it fails the 90% confidence level test. Under the scenarios of RCP4.5 and RCP8.5, the decreasing trend of the snow coverage rate is significant, passing the 99% confidence level test.