Temporal And Spatial Distribution Of Precipitation In The Data-scarce Area In The Hinterland Of The Qinghai-Tibet Plateau And Analysis Of Future Trend

This study investigates the precipitation dynamics in the hinterland of the Qinghai-Tibet Plateau,which is influenced by the interaction of westerly winds and the Indian Ocean monsoon,compounded by the region’s complex topography.Moreover,the meteorological station data in the study area is sparse and the station distribution is uneven.The applicability of the China Meteorological Forcing Dataset(CMFD)and Integrated Multi-satellit E Retrievals for GPM(IMERG)precipitation data for this region is initially evaluated by leveraging ground observation data.Subsequently,a spatio-temporal kriging method is employed to rectify the CMFD data based on ground station observations,facilitating the analysis of precipitation distribution patterns.Lastly,the study utilizes CN05.1 data to evaluate the ability of the global climate model data in CMIP6 to simulate precipitation in the Qinghai-Tibet Plateau’s hinterland.The paper also forecasts the average annual precipitation and the trend and spatial distribution of precipitation in four seasons in the future for 7 scenarios of 5 climate models in CMIP6.The results demonstrate that:(1)CMFD and IMERG accuracy comparison.On the daily scale,the R,Bias,RMSE,and MAE of the CMFD and the measured data are 0.82,-0.006,0.74,and 0.38,respectively,and the POD is greater than 0.6,the FAR is less than 0.45,the CSI is greater than 0.45,and the ETS is greater than 0.3.Furthermore,the R,Bias,RMSE,and MAE of the IMERG and the measured data are 0.69,0.02,1.07,and 0.53,respectively.It can be seen from the numerical accuracy indicators that CMFD performed better than IMERG precipitation products;On the monthly scale,both the scatter-point fitting line between precipitation products and the measured data is close to the 1:1 line.The deviations between IMERG and actual measurements in terms of the R,Bias,RMSE,MAE and CMFD indicators are 0.01,0.70,-1.67,-0.39,respectively.It can be concluded that IMERG performs better than CMFD based on these indicators;On a seasonal scale,the RMSE and MAE of CMFD and measured data in winter are slightly higher than those of IMERG data by 0.11 and 0.04 mm,respectively,and are better than IMERG precipitation products in other seasons.On the spatial scale,the spatial distribution of CMFD precipitation is generally consistent with the measured data,and the spatial performance of each evaluation index,CMFD data is better than IMERG data.(2)The spatio-temporal kriging correction of CMFD improves the R~2,RMSE,and R values relative to the uncorrected data for most stations,with the exception of Xiaozaohuo and Golmud stations.The RMSE values between corrected CMFD and observed data exhibit greater reduction in the central portion of the study area,and the spatial performance of RMSE is superior to the uncorrected data in most months.(3)The analysis of the revised CMFD data reveals that the maximum daily precipitation was recorded on July 11,2016,at 8.93 mm,while the maximum monthly precipitation recorded on June,2015,at 84.97 mm.Days with daily precipitation below 1 mm constituted 73.9%of the total number of days.Spatially,precipitation diminishes progressively from southeast to northeast,particularly during summer.Altitudes below 4250 m exhibit increasing daily precipitation,spring,summer,and autumn with rising altitude,whereas no discernible relationship exists at higher elevations.Winter daily precipitation primarily ranges between 0and 0.2 mm and does not exhibit a clear correlation with altitude.(4)The spatio-temporal analysis of CMIP6 data indicates minimal deviations between observed precipitation values in the historical period(1961-2014)and model-simulated values,exhibiting strong temporal-spatial correlation.Future projections suggest an overall increase in average annual precipitation,with substantial precipitation anomaly percentage increases in the SSP3-7.0 and SSP5-8.5 scenarios from 2021 to 2100.Areas with high precipitation anomaly percentages are predominantly situated at the source of the Lancang River in the southeast.Seasonal analyses reveal that the SSP3-7.0 scenario exhibits the most rapid precipitation increase in summer and winter,while the SSP5-8.5 scenario demonstrates the most rapid increase in spring and autumn.Precipitation increase is most pronounced in summer and least in winter,indicating marked seasonal and regional variations.