| In this paper,the regional climate model Reg CM4 is used to analyze the atmosphere with different resolutions(2.5°×2.5°,1.5°×1.5°,0.75°×0.75°)and design a one-way nested test(off-line driving of coarse resolution analog field Very high resolution test),carrying out a 10-year(1989-1999)numerical simulation of the Yangtze River Delta region to explore the effects of different resolution driving field data on the simulation spin-up time and the uncertainty of the simulation.Based on the research results,the 1.5°×1.5°driving field data was selected.Based on the spatial resolution of different modes(50km and 25km)and the cloud microphysical scheme(subex,Nogherotto/Tompkins,WSM5 scheme),the Changsan area was carried out for10 Year(2000-2009)simulation,Compare the results with the European Center ERA5data and the International Satellite Cloud Climatology Project ISCCP D2 dataset(2.5°×2.5°),and calculate the water vapor flux,depending on the heat source Q1and the water vapor sink Q2 to analyze the cause of the simulation error.According to the research and analysis,the following conclusions are drawn:(1)According to the test results of this article,the spin-up time is about 6 months after the start of the model,Compared with the coarse-resolution ERA40,selecting the higher-resolution initial drive fields EIN15 and EIN75 can make the analog value adjust faster and tend to the climate equilibrium,thereby effectively shortening the spin-up time.From the later simulation results,compared with the ERA40 test,the higher-resolution driving field has a smaller error after entering the climate equilibrium state;however,the test results of EIN15 and EIN75 are not different(the annual average temperature difference is 0.02?);Judging from this test,the overall simulation effect of FNEST is not ideal.Both the convergence of the early Spin-up stage and the later simulation results show a large instability.The current offline nesting method is in the20km high-resolution test There is also a certain systematic error.(2)The influence of different cloud microphysics schemes on the spatial and temporal distribution of cloud water presents a great difference.The subex scheme simulates the total cloud cover and precipitation in the case where the root mean square error is not different,and the spatial correlation The coefficient is higher,followed by the WSM5 program.The simulation error of the total cloud cover is mainly due to the inaccuracy of the spatial correlation simulation of low cloud cover by various schemes.In addition to the subexperiment under 50km,the errors of large-scale precipitation simulation account for more than 60%of the total precipitation simulation error.As far as the test results in this chapter are concerned,there is no obvious difference between the 50km and 25km test results of the subex and WSM5 schemes.Compared with the NT50 test,the NT25 test has an increase of 0.06 in the total cloud cover spatial correlation coefficient,and the error ratio in the simulation of large scale precipitation16%reduction.(3)Based on the analysis of water vapor flux and Q1Q2,the simulation error of the subex and WSM5 schemes for total cloud cover and precipitation is mainly derived from the overestimation of the water vapor flux in the lower troposphere over15g/(s*hpa*cm).In addition,the deviation of the simulation of the spatiotemporal characteristics of cloud water in the model mainly comes from the positive deviation of local Q1 exceeding 8W/m~2 and the positive deviation of Q2 exceeding 30W/m~2 in each experiment.The overestimation of total cloud cover and precipitation is mainly due to the vertical variation of Q1 and the local variation of Q2,but the improvement of the spatial resolution can significantly improve the simulation error of the local variation of Q2. |
PDF Full Text Request