Causes Of Precipitation Types Change And Forecast Of Future Trends In The Tianshan Mountains Area,China

In the context of global warming,the increase rate of average temperature in China is significantly higher than that of the average temperature in the northern hemisphere over the same period.Compared with northern and eastern China,the increase of asymmetric temperature in the northwest region is larger.And the precipitation types change from snowfall to rainfall.It is the same in the Tianshan Mountains area.The change of different precipitation types will inevitably affect the process of regional water resources production and the distribution during the year.Using the daily precipitation,average temperature,relative humidity,average pressure,average wind speed,sunshine hours,small evaporation,altitude,slope aspect,longitude and latitude data of 28 meteorological stations in Tianshan Mountain from 1960 to 2018,this paper separated the main precipitation types and calculated the rainfall days/rainfall days ratio(RPR)that characterizes the change in precipitation types,and constructed an indicator system of impact factors.Based on the artificial neural network,the decision-making and experimental evaluation laboratory method(BP-DEMATEL)and the multiple linear regression model(MLR)were used to identify the key driving factors for precipitation types changes in the Tianshan Mountains area.On this basis,the CMIP5 global climate model was used to evaluate the RPR simulation capability of the Tianshan Mountains area during 1961-2018.On this basis,the CMIP5 global climate model was used to evaluate the RPR simulation capabilities of the Tianshan Mountains from 1961 to 2018.Then this study estimated the future RPR of Tianshan area.The main conclusions obtained in this article are as follows:(1)From 1960 to 2017,the RPR of the Tianshan Mountains area showed a gradual growth trendwith two stages.The first stage was from the early 1960 s to the early 1990 s.And the second stage was from the early 1990 s to 2017.The average RPR of the two stages was 0.63 and 0.65,respectively.The change of RPR is altitude-dependent.The change of RPR with altitude also showed two stages: when the altitude is lower than 1200 m,the average RPR is 0.84 with relatively gentle change.when the altitude is over 1650 m,the average RPR is 0.47 with more dramatic change.In addition,the RPR changes were topographical.The average annual RPR of the south slope was 0.77,while the north slope was 0.57.The RPR of the south slope was significantly higher than the north slope.(2)Among all index factors,air temperature was the only strong driving factor in the BP-DEMATEL model and its contribution rate(lmg)was the largest in the MLR model,indicating that air temperature was the key driving factor affecting change of precipitation types in the Tianshan Mountains area.Land surface temperature,altitude,latitude,and air pressure had strong effects on other factors and played a driving role in the change of precipitation types,which can be considered as the secondary driving factors that affect the change of precipitation types.In addition to topographical factors and geographical factors,wind speed,evaporation and relative humidity are the result factors of RPR,which are closely related to other factors and are greatly affected by other factors,which indirectly.(3)The simulated value of CMIP5 model for air temperature was highly correlated with the observed value,while the simulated effect for precipitation was poor.Therefore,the precipitation data provided by CMIP5 was not used.Instead,the temperature data provided by CMIP5 and some meteorological factors were used to predicted RPR through the BP neural network.Using historical data to verified the simulation effect of the BP neural network model on RPR,the calibration and verification of 28 meteorological stations were Satisfactory and above,so the model could be used to further estimate the future monthly RPR.(4)Under the three scenarios,the RPR in the late autumn and early spring months in the Tianshan Mountains from 2010 to 2100 increased from the baseline period.The higher the emission scenario,the greater the RPR.The RPR value under RCP2.6,RCP4.5 and RCP8.5 scenarios will increase by 4.36%,8.27% and 12.36%,respectively.Under the same emission scenario,the later the time,the greater the RPR,and the more obvious the trend of snowfall into rainfall.In different months in the same station,RPR presented different changing characteristics.RPR at different stations in the same month showed different changing characteristics.Most of the future scenarios conform to the RCP emission scenario.the higher the RPR distribution,the greater the RPR distribution.There will be also cases where the RCP emission scenario is inconsistent but the range of RPR is consistent.(5)During the historical period,the increase rate of RPR showed that the northern slope was larger than the southern slope.The future spatial changes of increase rate of RPR will show a different trend from the present.The increase in RPR in the low-emission scenario in the initial stage show that the south slope will be greater than the north slope,while the increased in the south-north slope at the initial stage of the medium-emission scenario will be consistent.In terms of time change,as the year increased in the low-emission scenario,the more stations with no change trend,the change in tends of RPR to be stable.In the medium-emission scenario,RPR will experienced an increase in the entire Tianshan,and the station on the northern slope eventually stabilized.Under the high-emission scenario,the entire Tianshan Mountain will show a significant increase in all time periods.