Issue with Land Use Timeseries and Surface Data Processing for CLM5.0

[Will Smith]

What version of the code are you using?
release-cesm2.2.0

Have you made any changes to files in the source tree?
No

Describe your problem or question:
Hi all,

I am currently using CLM5.0 along with the emission_ssp585 dataset to generate land use timeseries data and surface data for my
customised grid. I have downloaded all the necessary input data, but I consistently encounter an error when processing the year after 2016.
The error looks like:

```
Program aborted. Backtrace:
#0 0x40225f in ???
#1 0x406123 in ???
#2 0x42b4c5 in ???
#3 0x442ad1 in ???
#4 0x40207c in ???
#5 0x2af7f5b89554 in ???
#6 0x4020ac in ???
#7 0xffffffffffffffff in ???
sh: line 1: 83768 Aborted /mnt/iusers01/fatpou01/sees01/s29826zs/CESM/my_cesm_sandbox_2.2.0/components/clm/tools/mksurfdata_map/mksurfdata_map < surfdata_ne0np4.UK_ne30x16_SSP5-8.5_78pfts_CMIP6_simyr1850_c250203.namelist
ERROR in mksurfdata_map: 34304
```

The command I used:
./mksurfdata.pl -years 1850-2100,1850,2000 -ssp_rcp SSP5-8.5 ${CROPSTRING} -res usrspec -usr_gname ${VRgridName} -usr_gdate ${cdate} -usr_mapdir ${TMPDIR} -l ${CSMDATA_PATH} -exedir ${CESMROOT}/components/clm/tools/mksurfdata_map

I have attached the log files detailing the issue. Would you be able to provide any insight into what might be causing this problem?

Any guidance would be greatly appreciated.
Will

 

[slevis]

I have a non-ideal suggestion, but it's a way of troubleshooting at least:
I would try making separate fdyndat files using separate mksrf_fdynuse files, one for 1850-2015 and the other for 2016-2100. If that works, then you may find a way to merge the two fdyndat files into one. Or you could perform the two time periods as separate simulations.

 

[Will Smith]

I have a non-ideal suggestion, but it's a way of troubleshooting at least:
I would try making separate fdyndat files using separate mksrf_fdynuse files, one for 1850-2015 and the other for 2016-2100. If that works, then you may find a way to merge the two fdyndat files into one. Or you could perform the two time periods as separate simulations.

Hi Slevis,

Thank you very much for your suggestions.

I have encountered an issue while generating the fdyndat files for the period 2016-2100. When processing the files for 1850-2015, everything worked without any problems. However, for 2016-2100, I used the following command:

```
./mksurfdata.pl -years 2016-2100 -ssp_rcp SSP5-8.5 ${CROPSTRING} -res usrspec -usr_gname ${VRgridName} -usr_gdate ${cdate} -usr_mapdir ${TMPDIR} -l ${CSMDATA_PATH} -exedir ${CESMROOT}/components/clm/tools/mksurfdata_map
```
My intention was to use the SSP5-8.5 emissions. However, when running the command, I encountered an error (please find the log attached). The log suggests that I am missing files for SSP1-2.6, despite explicitly specifying SSP5-8.5 in the command.

Is there something I might have done incorrectly? Any guidance would be appreciated.

 

[oleson]

That is interesting. I tried the following for the standard resolution of 0.9x1.25 and SSP5-8.5 with release-cesm2.2.0:

mksurfdata.pl -ssp_rcp SSP5-8.5 -res 0.9x1.25 -years 1850-2100

and it seemed to work fine.
The last couple of lines of the landuse_timeseries_SSP5-8.5_78pfts_CMIP6_simyr1850-2100.txt file that is generated are:

/glade/derecho/scratch/oleson/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2099.c171005.nc
/glade/derecho/scratch/oleson/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2100.c171005.nc

I guess your lines must have SSP1RCP26?
You could try the command above to start and see if it works for you. Maybe there is a problem with the namelist in your version.

 

[Will Smith]

That is interesting. I tried the following for the standard resolution of 0.9x1.25 and SSP5-8.5 with release-cesm2.2.0:

mksurfdata.pl -ssp_rcp SSP5-8.5 -res 0.9x1.25 -years 1850-2100

and it seemed to work fine.
The last couple of lines of the landuse_timeseries_SSP5-8.5_78pfts_CMIP6_simyr1850-2100.txt file that is generated are:

/glade/derecho/scratch/oleson/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2099.c171005.nc
/glade/derecho/scratch/oleson/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2100.c171005.nc

I guess your lines must have SSP1RCP26?
You could try the command above to start and see if it works for you. Maybe there is a problem with the namelist in your version.

Hi,

Thank you very much for your response.
The last couple of lines in the landuse_timeseries_SSP5-8.5_78pfts_CMIP6_simyr2000-2100.txt file of mine that is generated are as follows:

```
/mnt/iusers01/fatpou01/sees01/s29826zs/scratch/Projects/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2099.c171005.nc 2099
/mnt/iusers01/fatpou01/sees01/s29826zs/scratch/Projects/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2099.c171005.nc 2099
/mnt/iusers01/fatpou01/sees01/s29826zs/scratch/Projects/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2100.c171005.nc 2100
/mnt/iusers01/fatpou01/sees01/s29826zs/scratch/Projects/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2100.c171005.nc 2100
```
There is no occurrence of SSP1RCP26. I have attached the landuse_timeseries_SSP5-8.5_78pfts_CMIP6_simyr2000-2100.txt file for your reference. I am currently testing the following command as you suggested to check whether there is an issue: mksurfdata.pl -ssp_rcp SSP5-8.5 -res 0.9x1.25 -years 1850-2100

Many thanks.

 

[Will Smith]

Hi all,

A quick update. When I used SSP370 data and switched the ESMF version to 8.6.1 (previously 8.6.0, which also caused issues with SSP370), there were no problems, and the run completed successfully. However, when applying the same ESMF version to SSP585, the error changed to the following:

```
Program aborted. Backtrace:
#0 0x40225f in ???
#1 0x406123 in ???
#2 0x42b4c5 in ???
#3 0x43d7a8 in ???
#4 0x40207c in ???
#5 0x2ad56d578554 in ???
#6 0x4020ac in ???
#7 0xffffffffffffffff in ???
sh: line 1: 39378 Aborted /mnt/iusers01/fatpou01/sees01/s29826zs/CESM/my_cesm_sandbox_2.2.0/components/clm/tools/mksurfdata_map/mksurfdata_map < surfdata_ne0np4.UK_ne30x16_SSP5-8.5_78pfts_CMIP6_simyr2016_c250206.namelist
ERROR in mksurfdata_map: 34304
```
I am wondering whether this could be an issue with the ESMF version or with surfdata_ne0np4.UK_ne30x16_SSP5-8.5_78pfts_CMIP6_simyr2016_c250206.namelist? Attached is the log files, ssp585, ssp370 namelists.

 

[oleson]

If it works with SSP370 but not SSP585 I doubt it would have anything to do with the ESMF version. Is there a PET log in that directory that might provide more information?

Since it is stopping at making pfts for 2016, according to the log, maybe make sure this file is a valid netcdf file, not corrupted in some way:

/mnt/iusers01/fatpou01/sees01/s29826zs/scratch/Projects/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2016.c171005.nc

 

[Will Smith]

If it works with SSP370 but not SSP585 I doubt it would have anything to do with the ESMF version. Is there a PET log in that directory that might provide more information?

Since it is stopping at making pfts for 2016, according to the log, maybe make sure this file is a valid netcdf file, not corrupted in some way:

/mnt/iusers01/fatpou01/sees01/s29826zs/scratch/Projects/inputdata/lnd/clm2/rawdata/pftcftdynharv.0.25x0.25.SSP5-8.5.simyr2016-2100.c171005/mksrf_landuse_SSP5RCP85_clm5_2016.c171005.nc

Hi,

Thank you very much for your suggestions. I believe I have identified the issue—it appears to be related to the file mksrf_landuse_SSP5RCP85_clm5_2016.c171005.nc. When I bypass this file, it runs successfully for the years 2017–2025.

I think I will need to re-download these datasets to verify them properly. I had suspected many other causes but overlooked the data itself!

Many thanks,
Will