基于多种统计降尺度方法的未来降水预估研究——以青藏高原为例

Future Precipitation Projection Based on Multiple Statistical Downscaling Methods——A Case Study of Tibetan Plateau

虽然第六次耦合模式比较计划(Coupled Model Intercomparison Project 6,CMIP6)能很好地预测大尺度气候要素,但是其在预测流域尺度方面的效果与实测数据仍有差别,尤其是在青藏高原这种高海拔、地形复杂地区,气候模式所产生的误差更大。基于最新一代高分辨率CMIP6模式历史情景和SSP126、SSP245、SSP370、SSP585等多种未来气候排放情景,研究使用包括偏差校正、KNN、SDSM等多种统计降尺度方法进行降尺度分析,并对各自的预测性能进行了评估,在此基础上使用性能最佳的统计降尺度方式预估青藏高原地区的未来降水,对最终得到的预估降水的时空演变特征进行了详细的分析,并与青藏高原的历史降水情况进行了对比。结果表明,3种统计降尺度在青藏高原的适用性差异较大,线性回归降尺度方法的性能最佳,其次为偏差校正方法,最差为KNN类比方法。从未来降水预估情况分析,青藏高原未来80 a平均降水、降水极值等总体呈上升趋势但上升幅度较小,且空间分布情况变化不大。研究结果可为青藏高原水资源评价及规划与管理提供科学依据。

Although the Coupled Model Intercomparison Project 6(CMIP6)can well predict large-scale climatic factors,its effect on projecting watershed scales is still different from the measured data.The error of climate models is even bigger over the Tibetan Plateau,which is a high-altitude region with complicated terrain.Based on the historical scenario of the latest generation of high-resolution CMIP6 model and a variety of future climate emission scenarios such as SSP126,SSP245,SSP370,and SSP585,this paper conducts downscaling analysis and evaluates the projection performance of various statistical downscaling methods such as bias correction,KNN,and SDSM.On this basis,the best statistical downscaling method is used to project future precipitation over the Tibetan Plateau,and the spatial-temporal evolution characteristics of the projected precipitation are analyzed and compared with the historical precipitation over the Tibetan Plateau.The results reveal that the applicability amongst the three statistical downscaling methods in the Tibetan Plateau is large,with the linear regression downscaling method performing the best,followed by the bias correction method and the KNN analogy method.According to the analysis of future precipitation projections,the average precipitation and extreme precipitation over the Tibetan Plateau in the next 80 years will exhibit an overall upward trend,although the rise will be slight,and the spatial distribution will not change much.The results can provide a scientific foundation for the evaluation,planning,and management of water resources on the Tibetan Plateau.