Simulation and prediction of climatic elements in Jinghe River basin using downscaling method and CMIP5

Abstract: Under the background of frequent anthropogenic activities and global environment change, it is necessary to carry out the study of future climate change, for understanding of the characteristics of temporal and spatial evolution of future climate is of great significance to agricultural production and local water resources planning and management. Located in the middle of the Loess Plateau, China, the Jinhe River basin, the second largest river in Guanzhong area, is the major base of grain production in Shaanxi Province, which has flat land and developed agriculture since ancient times. With the establishment of the Guanzhong-Tianshui economic zone, which is a national key economic development zone and will greatly promote the rapid development of the economy in the whole western region, the security of water resources in Jinhe River basin is becoming more and more important for the economic development and local peace and stability. Therefore, in this paper, we took Jinhe River basin as the research object. Then the variation characteristics of rainfall and temperature in the future were focused and analyzed by using downscaling method and CMIP5 (coupled model intercomparison project 5) global climate models, which were applied to simulate and predict climatic elements in Jinghe River basin. There were 3 steps to establish the statistical downscaling models: firstly, determinate forecast factors based on the relationship between the factors and rainfall as well as the air temperature; then normalize the forecasting factors in the NCEP (national centers for environmental prediction) reanalysis data; eventually based on the reanalysis data and historical meteorological observation data of monthly precipitation and temperature from 7 weather stations in or around the basin from 1960 to 2010, the models were set up through screening major factors by using stepwise regression. In order to evaluate the effect of the model, the fitting degree R2, the relative error of the mean Rmean and the relative error of standard deviation Rsd were selected as the evaluation index. These models were calibrated so that the results could be satisfactory. Then, on the foundation of the calibrated models, future precipitation and temperature in the Jinhe River basin were predicted by atmospheric circulation factors came from CMIP5 experiments: the RCP8.5 (highest emission of greenhouse gases) and RCP4.5 (median emission of greenhouse gases) forcing pathways under CNRM-CM5. Finally, we analyzed the characteristics of temporal and spatial of future precipitation and temperature in the basin. The results show that: (1) The statistical downscaling models function better in simulating temperature than precipitation with higher fitting degree and lower relative error in average and standard deviation; (2) Precipitation in the basin from 2011 to 2050 showed a decreasing trend from south to north direction, and would increase only in winter with different magnitudes under 2 scenarios; moreover, the annual average precipitation was 356.41 mm under the RCP8.5 scenario, which was lower than RCP4.5 forcing pathway (374.19 mm); (3) The predicted temperature in the future was much higher under RCP8.5 scenario (9.32℃) compared to RCP4.5 scenario (8.96℃) because of greenhouse gas emission with different concentration; spatial distribution of future temperature in Jinghe River basin was characterized by higher in the north and west; besides, temperature in the basin would decrease in the late winter and early spring, but significantly increase in summer, and it was inferred that there was the possibility of occurrence of extremely high temperature in summer and extremely low temperature in winter. In conclusion, there is not only a trend of decreasing in precipitation, but also a risk of extreme weather events in Jinghe River basin, which should be paid attention to in the future water resources management and planning.