Analysis Of Intraseasonal Characteristics Of East Asian Precipitation In GRIST Model

At present,the research of integrated models more focuses on the long-term mean climate change,while the understanding of its intraseasonal timescale is relatively less.The research on precipitation characteristics of the model on the season timescale can not only help to understand the basic characteristics of the model effectively,but also greatly help to model development and error correction.Based on the seasonal climate hindcast experiment of the Global-to-Regional Integrated forecast Sys Tem(GRIST)model,the simulated intraseasonal variation of East Asian summer precipitation is analyzed.Compared with the results of observed data,i.e.,the China merged hourly precipitation analysis(CMPA)dataset and the Global Precipitation Measurement(GPM)dataset,and two global climate models,i.e.,the Community Atmosphere Model version 5(CAM5)and the superparameterized CAM5(SPCAM5),this study explores the inter-model differences and their influencing factors are explored.Then,the sensitivity of GRIST model with different grid resolutions(120,80,and 110-35 km)over East Asia and different experiments(continuous integration and seasonal climate hindcast)to the intraseasonal features of East Asian summer precipitation is further analyzed.The main conclusions are summarized as follows:There are great differences in the distribution of East Asian summer mean precipitation among different models.The spatial correlation coefficient between the GRIST model and the observation is higher,and it can basically reproduce 10-day mean precipitation variation in East China.Precipitation evolution in eastern China simulated by GRIST is basically consistent with observation,but with high rainfall.CAM5 is insensitive to the variation of 10-day mean precipitation variation and the simulated main flood season is earlier.Further,the analysis shows that the daily evolution of the sub-analytic scale source and sink terms and other parameters corresponded to the 10-day mean precipitation evolution simulated by the model,which reflects the direct influence of source and sink terms on the model and the intraseasonal variation of the model precipitation.GRIST can reproduce the position of rain belt and subtropical ridge,and their northward advancing process.It can also capture the northward characteristics of intrasesonal oscillation(ISO).In terms of the diurnal variation characteristics of precipitation,the simulation results of GRIST and SPCAM5 are roughly similar,but the peak time of precipitation over the Tibetan Plateau simulated earlier by GRIST,and it does not capture the eastward transmission signal of precipitation on the lee side of the plateau in mid-late July.There is no significant difference in the distribution of East Asian summer mean precipitation simulated by models with different grid resolutions and different experiments,and the spatial correlation coefficient between the model and the observation is above 0.72.Among them,the correlation of GRIST model formed by continuous integration test with 80-km grid resolution is the highest.As the gird resolution increasing,the variation of 10-day mean precipitation is strengthened and the rainfall in August is advanced and increases.There are also differences among different integration tests in the peak period of August.There is little difference in the location of the subtropical high ridge and the rainband simulated by different grid resolution models,while the location of the northwest Pacific subtropical high ridge and the rainband simulated by the seasonal climate hindcast experiment of GRIST model is southerly in June.In terms of diurnal variation characteristics,with the improvement of resolution,the peak time of precipitation in Southwest China and India is advanced,and the eastward transmission signal of precipitation is enhanced.