error message with new cami file

[nadinsel@uchicago_edu]

Hi all,I am trying to run deep-time simulations with CESM and running into problems creating a new cami file. I started with the original cami_create_cesm.csh and cesm_cami_create.ncl. I had to modify those files to include staggered latitudes. So now I am able to create a new cami.nc file. I used the original topography (USGS_gtopo30_0.47x0.63_remap_c061106) and skeleton file (cami_1980_01-01_0.47x0.63_L26_c071226). The newly created cami.nc file looks correct when looking at it with ncview.However, when I am running a modern 0.47x0.63 cam stand-alone configuration  with the new cami file (the cami input file is the only thing I changed) I get the following error message in the ccsm.log file:  *** ZM_CONV: IENTROPY: Failed and about to exit, info follows ****ZM_CONV: IENTROPY. Details: call#,lchnk,icol= 24212  11 lat: -41.59 lon: 168.75 P(mb)=  53.11 Tfg(K)= 206.51 qt(g/kg) =   0.27 qsat(g/kg) =    NaN, s(J/kg) =    NaNENDRUN:**** ZM_CONV IENTROPY: Tmix did not converge ****application called MPI_Abort(MPI_COMM_WORLD, 1) - process 137*** ZM_CONV: IENTROPY: Failed and about to exit, info follows ****ZM_CONV: IENTROPY. Details: call#,lchnk,icol= 2****   2 lat:  42.53 lon:  17.50 P(mb)=  37.23 Tfg(K)= 195.98 qt(g/kg) =   0.17 qsat(g/kg) =    NaN, s(J/kg) =    NaNENDRUN:**** ZM_CONV IENTROPY: Tmix did not converge ****application called MPI_Abort(MPI_COMM_WORLD, 1) - process 376*** ZM_CONV: IENTROPY: Failed and about to exit, info follows ****ZM_CONV: IENTROPY. Details: call#,lchnk,icol= 24105  11 lat: -43.00 lon: 169.38 P(mb)=  53.11 Tfg(K)= 206.56 qt(g/kg) =   0.26 qsat(g/kg) =    NaN, s(J/kg) =    NaNENDRUN:**** ZM_CONV IENTROPY: Tmix did not converge ****application called MPI_Abort(MPI_COMM_WORLD, 1) - process 133APPLICATION TERMINATED WITH THE EXIT STRING: Hangup (signal 1) I am wondering if anyone came across the same problem or an error message like that or if you have any ideas how to solve that issue?Thanks,Nadja

 

[nanr]

Nadja - It sounds as though you have bad values in your cami file.  Have you checked that there are no missing values?  Especially near the poles?

 

[nadinsel@uchicago_edu]

Hi Nan,I checked the nc file with ncdump and grep and couldn't find any missing values. However, here might be the issue:In cesm_cami_create.ncl nlon and nlat are defined as dimsizes (lon) and dimsizes (lat), thus nlat is 384. The thermal wind in my skeleton file has staggered latitudes with nlat =383. Therefore, when cesm_cami_create.ncl tries to calculate thermal winds       u(:,:,0,:) = u(:,:,1,:)      u(:,:,nlat-1,:) = u(:,:,nlat-2,:)the second command line is impossible. I tried to fix the problem by commenting out that line and writing the output thermal winds with staggered latitudes so it matches the original cami file used for modern half-degree resolution runs. But maybe that becomes an issue when replacing the cami file? If that's the case, how can I calculate the thermal wind in the cesm_cami_create.ncl file, when my original U is actually a Ustaggered in the skeleton file? 

 

[nanr]

Nadja - Try using these updated cami creation scripts and see if that helps.  You will need a default cesm1.x "cam.i" file as a template. 

 

[nadinsel@uchicago_edu]

Hi Nan,thanks for the updated files. They work fine to create a new cami file. However, using the newly-created cami file for a CESM run results in the same error message: *** ZM_CONV: IENTROPY: Failed and about to exit, info follows ****ZM_CONV: IENTROPY. Details: call#,lchnk,icol= 28431   3 lat:  12.92 lon: 128.13 P(mb)=  37.23 Tfg(K)= 194.92 qt(g/kg) =   0.20 qsat(g/kg) =    NaN, s(J/kg) =    NaN ENDRUN:**** ZM_CONV IENTROPY: Tmix did not converge **** So what I found is that the model can't simulate convection in a 0.5 degree resolution when the temperature profil lacks a vertical gradient. The cami creation scripts all ascribe a temperature profile with a temperature gradient from the equator to the poles, but the same equator - pole profile exists for all atmospheric pressures. When I introduce a vertical temperature profile in the cami file the error message disappears and the model runs fine. So my question is: What is the best vertical temperature profile to start with for the equator and poles with atmospheric pressures from 993 to 3.5 mb?