nabicherry@gmail_com
New Member
I tested 2 experiments CESM1.0.4 using FAMIP case (1979 – 2005 year run) with f19_f19 resolution. One experiment (Ex1) is performed with default setting.The other experiment (Ex2) is performed with changing PCT_PFT in file ‘surfdata.pftdyn_1.9x2.5_simyr1850-2005_c091108.nc’ In Ex2, the variable PCT_PFT is fixed from 1979 to 2005 by value of 1979. Thus PCT PFT does not change from 1979 to 2005 in Ex2.Except PCT_PFT, other boundary and initial condition is same for Ex1 and Ex2. I think Ex1 and Ex2 are performed with same PCT_PFT at the year of 1979, thus they show same results at 1979.However, Ex2 is differ from Ex1 even at first step of integration. Figure in attached file shows difference of surface temperature between Ex1 and Ex2.The surface temperature is different at 8 points only, and the temperature difference is amplified.Otherwise, leaf area index (LAI) is same at first step. After second step, LAI is changed by the temperature difference. I tested additional experiments with changing PCT_PFT.In all experiment, Then, surface temperature of first step is same with Ex1 or Ex2.Generally, I change PCT_PFT only one year (ex: 1980, 1981, 1990, 2004) then the result is same with Ex1.If I changed PCT_PFT more than two years (ex: 1980-1981, 1979-2005, 1980-2005) then the result is same with Ex2. I don’t know why the model produce different results.Could you help me? I want to two experiments (Ex1 and Ex2) produce same result at the year of 1979. For model simulation, I use intel fortran compiler version v. 11, netcdf v 3.6.3, mvapich v. 1.4,I think system is not source of this problem.