Characteristics Of Climate Change In The Qilian Mountains Over The Last 40 Years And The Large-scale Circulation

The Qilian Mountains is a hotspot region in global climate changing research,but lacking meteorological stations,especially in high-altitude mountainous regions,which has limited the in-depth study of climate change in the Qilian Mountains.Therefore,it is essential to examine the robustness of the reanalysis data.The atmospheric circulation is a crucial influencing factor of global climate change,the investigation of the large-scale circulation and climate changeis meaningful under the warming background.In this study,2m temperature,pressure,wind speed and precipitation from 19 automaticmeteorological stations during the period 1980-2019,were used to evaluate the applicability of four global reanalysis products(ERA5,ERA-Interim,JRA-55 and HAR)by comparing the correlation coefficient(R),mean bias(BIAS),root mean square error(RMSE)and mean absolute error(MAE).ERA5,ERA-Interim,and JRA-55 reanalysis datasets were used to analyze the climate change characteristics of the Qilian Mountains region in the past 40 years,and the circulation background of climate change in the Qilian Mountains region was analyzed by the reanalyses and the northern hemisphere remote correlation indices(Arctic Oscillation,AO).The main conclusions are as followed.(1)For temperature and pressure,the correlations of all four reanalysis datasets and meteorological stations are relatively high,while the correlations of wind speed are slightly worse and precipitation is the worst.The performance of the reanalysis datasets of different meteorological elements varies,with ERA5 performing best for air temperature,HAR and ERA5 for pressure,ERA5 for wind speed,and ERA-Interim for precipitation at weather stations,while ERA5 is better at simulating abundant precipitation at high altitudes.In terms of spatial distribution,the simulation skill of temperature,pressure and wind speed is poor in regions with large elevation difference in reanalyses,while the simulation skill is better in regions with flat terrain.For precipitation,the error is larger in the eastern part of Qilian Mountains,where the precipitation is large than others.(2)For ERA5,ERA-Interimand JRA-55,there is basically a similar inter-decadal variation in temperature,with a general trend of increasing temperature and a faster warming rate in the Qilian Mountains than the Tibetan Plateau from 1980 to 2018.All three reanalyses show a rising trend of total annual precipitation and the trend are faster than the national average precipitation growth rate,but JRA-55 shows an anomalous and rapid increase from 2015 to 2019.(3)The correlations of mean temperature and total precipitation with AO in the Qilian Mountains simulated by the three reanalysis datasets have very similar performance.For the correlation coefficients of annual mean temperature and winter mean temperature with AO are basically spatially distributed with a higher correlation coefficient in the north and a lower correlation coefficient in the south,and the correlation coefficient in winter is slightly higher than the annual mean.The correlation between total precipitation and total summer precipitation with AO in Qilian Mountains are basically positive,and the correlation between total precipitation and AO in summer is stronger than that of total annual precipitation;the correlation coefficient between total precipitation and AO in winter shows a distribution characteristic of low values in the northwest and high values in the southeast.Between 1958 and 2019,the annual mean 500 h Pa geopotential height field in the Qilian Mountains shows a fluctuating upward trend.The annual mean 500 h Pa geopotential height field increases significantly around 1980,basically the spatial distribution pattern of higher altitudes with less variation and lower altitudes with more variation is observed.