ERROR: GLACIER_REGION NOT on surfdata file

[dayon]

Dear Scientists

In order to simulate the regional runoff and streamflow, I cut the domain file and surface data file based on provide 0.125x0.125 domains and surfdata_map by using the clm tools "getregional_datasets.pl"
the provided orignal files are as follows:
domain.lnd.0.125x0.125_tx0.1v2.140704.nc
surfdata_0.125x0.125_simyr2000_c150114.nc

I cut the region with the 102E-108E, 23-28N. The domain and surface nc file is generated successful!

However, when I run the CESM2.1.3 with the compset named "I2000Clm50SpGs", which is descriped "SGLC Stub glacier (land ice) component", which means that the glacier should be stubbed. But when typing "case.submit", Errors occurs

the error log shows that
ERROR: ERROR: GLACIER_REGION NOT on surfdata file
ERROR in /home/dayon/cesm2_1_3/components/clm/src/main/glcBehaviorMod.F90 at line 662

It is so strange, the glacier is stubbed, why the surface file need GLACIER_REGION?

Any suggestion is appreciated. Thanks a lot.

 

[sacks]

GLACIER_REGION is needed even when running with a stub GLC model: these regions determine the values for a few model variables that are needed no matter whether you're running with CISM or a stub GLC model.

For your small region, you can safely create a GLACIER_REGION field that is 0 in every grid cell.

 

[dayon]

Hi, Scaks. I create a GLACIER_REGION field that is 0 in every grid cell. but it still showed error:

ERROR: ERROR in /home/dayon/cesm2_1_3/components/clm/src/main/surfrdUtilsMod.F90 at line 70
sum is: 1.1576044285695568E-312

 

[sacks]

I am moving this to the land forums. Please provide more information, including more information from the log file (I think it should show the name of the variable in error before printing the sum) and more details on the steps you took that led up to this error.

 

[dayon]

Hi, Sacks. The error log is as attached.

 

[sacks]

I don't see an attachment. However, reading back through your original comment, I think I see the problem: you are subsetting an older surface dataset that is not compatible with the more recent model code. I'm going to reach out to others to see what is generally recommended in this situation.

 

[sacks]

I just see that you had already posted the same question to the land forums (ERROR: GLACIER_REGION NOT on surfdata file) where it was answered by Keith Oleson. So I consider this resolved. In the future, please try not to double-post questions, since this leads to multiple people trying to answer the same question.