Dear Ming Chen, Sorry to reply so late. I thought my simulation result have some problems, cause the RAINC in my grids is NAN. Then I checked there have some values in other grids. But there is still a problem. Especially for precipitation, my simulated T2 and observation temperature correlation coefficient can only reach 0.75. My CMAQ simulations are not good, so I wonder if it is related to the meteorological results. I want to simulate a strong convective process. So how can I get a better performance in my simulation? My domain is PRD, the May, and June of 2015. Can you give me some help? I've uploaded my namelist.input and input.wps file. Thanks a lot.By saying RAINC is all null, do you mean it is NAN?