Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Empty_ container

Yonas

Yonas
New Member
Hello, i am trying to use the CESM container:escomp/cesm-2.2 without the Jupyter notebook interface. The one with Jupyter notebook is working well in my laptop but when i try to pull and run the one without notebook i am getting empty container. Here is the code i used to pull and run:

docker pull escomp/cesm-2.2:latest
docker run -t -d --name latest escomp/cesm-2.2:latest
docker exec -it latest bash

when i list the files inside the container, i found these files:

[user@cesm2.2 ~]$ ls -a

. .. .bash_history .bash_logout .bash_profile .bashrc .emacs

can you please help me how to figure out this issue?
 

ryanj

Ryan Johnson
ISG
Staff member
I think that directory is empty by default and is where you can save any files or input data via a mounted directory from your local machine. More information in the "Quick Start" section from the link below:

CESM is installed in /opt/ncar/cesm2 inside of the container
Bash:
[user@cesm2.2 ~]$ ls -al /opt/ncar/cesm2/
total 2220
drwxr-xr-x 1 root root    4096 Nov  1 23:10 .
drwxr-xr-x 1 root root    4096 Nov  1 23:09 ..
drwxr-xr-x 8 root root    4096 Nov  1 23:09 .git
drwxr-xr-x 3 root root    4096 Nov  1 23:09 .github
-rw-r--r-- 1 root root     263 Nov  1 23:09 .gitignore
-rw-r--r-- 1 root root    7471 Nov  1 23:09 CODE_OF_CONDUCT.md
-rw-r--r-- 1 root root 2184627 Nov  1 23:09 ChangeLog
-rw-r--r-- 1 root root    2433 Nov  1 23:09 ChangeLog_template
-rw-r--r-- 1 root root    1842 Nov  1 23:09 Externals.cfg
-rw-r--r-- 1 root root     182 Nov  1 23:09 Externals_cime.cfg
-rw-r--r-- 1 root root    5498 Nov  1 23:09 LICENSE.txt
-rw-r--r-- 1 root root    8896 Nov  1 23:09 README.rst
drwxr-xr-x 1 root root    4096 Nov  1 23:10 cime
drwxr-xr-x 1 root root    4096 Nov  1 23:11 cime_config
drwxr-xr-x 1 root root    4096 Nov  1 23:10 components
-rwxr-xr-x 1 root root    2475 Nov  1 23:09 describe_version
drwxr-xr-x 3 root root    4096 Nov  1 23:09 doc
drwxr-xr-x 5 root root    4096 Nov  1 23:09 manage_externals


I also followed your steps in creating / running the container and was able to successfully run the test case from the quick start
Code:
[user@cesm2.2 ~]$ create_newcase --case test1 --compset F2000climo --res f19_g17 --run-unsupported
Compset longname is 2000_CAM60_CLM50%SP_CICE%PRES_DOCN%DOM_MOSART_CISM2%NOEVOLVE_SWAV
Compset specification file is /opt/ncar/cesm2/components/cam//cime_config/config_compsets.xml
Automatically adding SIAC to compset
Automatically adding SESP to compset
Compset forcing is 1972-2004
ATM component is CAM cam6 physics:
LND component is clm5.0:Satellite phenology:
ICE component is Sea ICE (cice) model version 5 :prescribed cice
OCN component is DOCN   prescribed ocean mode
ROF component is MOSART: MOdel for Scale Adaptive River Transport
GLC component is cism2 (default, higher-order, can run in parallel):cism ice evolution turned off (this is the standard configuration unless you're explicitly interested in ice evolution):
WAV component is Stub wave component
IAC component is Stub iac component
ESP component is Stub external system processing (ESP) component
Pes     specification file is /opt/ncar/cesm2/components/cam//cime_config/config_pes.xml
Compset specific settings: name is SSTICE_DATA_FILENAME and value is $DIN_LOC_ROOT/atm/cam/sst/sst_HadOIBl_bc_1x1_2000climo_c180511.nc
Could not find machine match for 'de9bc9acbce1' or 'de9bc9acbce1'
Machine is container
Pes setting: machine match is container
Pes setting: grid          is a%1.9x2.5_l%1.9x2.5_oi%gx1v7_r%r05_g%gland4_w%null_z%null_m%gx1v7
Pes setting: compset       is 2000_CAM60_CLM50%SP_CICE%PRES_DOCN%DOM_MOSART_CISM2%NOEVOLVE_SWAV_SIAC_SESP
Pes setting: tasks       is {'NTASKS_ATM': -1, 'NTASKS_LND': -1, 'NTASKS_ROF': -1, 'NTASKS_ICE': -1, 'NTASKS_OCN': -1, 'NTASKS_GLC': -1, 'NTASKS_WAV': -1, 'NTASKS_CPL': -1}
Pes setting: threads     is {'NTHRDS_ATM': 1, 'NTHRDS_LND': 1, 'NTHRDS_ROF': 1, 'NTHRDS_ICE': 1, 'NTHRDS_OCN': 1, 'NTHRDS_GLC': 1, 'NTHRDS_WAV': 1, 'NTHRDS_CPL': 1}
Pes setting: rootpe      is {'ROOTPE_ATM': 0, 'ROOTPE_LND': 0, 'ROOTPE_ROF': 0, 'ROOTPE_ICE': 0, 'ROOTPE_OCN': 0, 'ROOTPE_GLC': 0, 'ROOTPE_WAV': 0, 'ROOTPE_CPL': 0}
Pes setting: pstrid      is {}
Pes other settings: {}
 Compset is: 2000_CAM60_CLM50%SP_CICE%PRES_DOCN%DOM_MOSART_CISM2%NOEVOLVE_SWAV_SIAC_SESP
 Grid is: a%1.9x2.5_l%1.9x2.5_oi%gx1v7_r%r05_g%gland4_w%null_z%null_m%gx1v7
 Components in compset are: ['cam', 'clm', 'cice', 'docn', 'mosart', 'cism', 'swav', 'siac', 'sesp']
No project info available
No charge_account info available, using value from PROJECT
cesm model version found: cesm2.2.0
Batch_system_type is none
 Creating Case directory /home/user/test1
[user@cesm2.2 ~]$ ls
test1
 

Yonas

Yonas
New Member
I think that directory is empty by default and is where you can save any files or input data via a mounted directory from your local machine. More information in the "Quick Start" section from the link below:

CESM is installed in /opt/ncar/cesm2 inside of the container
Bash:
[user@cesm2.2 ~]$ ls -al /opt/ncar/cesm2/
total 2220
drwxr-xr-x 1 root root    4096 Nov  1 23:10 .
drwxr-xr-x 1 root root    4096 Nov  1 23:09 ..
drwxr-xr-x 8 root root    4096 Nov  1 23:09 .git
drwxr-xr-x 3 root root    4096 Nov  1 23:09 .github
-rw-r--r-- 1 root root     263 Nov  1 23:09 .gitignore
-rw-r--r-- 1 root root    7471 Nov  1 23:09 CODE_OF_CONDUCT.md
-rw-r--r-- 1 root root 2184627 Nov  1 23:09 ChangeLog
-rw-r--r-- 1 root root    2433 Nov  1 23:09 ChangeLog_template
-rw-r--r-- 1 root root    1842 Nov  1 23:09 Externals.cfg
-rw-r--r-- 1 root root     182 Nov  1 23:09 Externals_cime.cfg
-rw-r--r-- 1 root root    5498 Nov  1 23:09 LICENSE.txt
-rw-r--r-- 1 root root    8896 Nov  1 23:09 README.rst
drwxr-xr-x 1 root root    4096 Nov  1 23:10 cime
drwxr-xr-x 1 root root    4096 Nov  1 23:11 cime_config
drwxr-xr-x 1 root root    4096 Nov  1 23:10 components
-rwxr-xr-x 1 root root    2475 Nov  1 23:09 describe_version
drwxr-xr-x 3 root root    4096 Nov  1 23:09 doc
drwxr-xr-x 5 root root    4096 Nov  1 23:09 manage_externals


I also followed your steps in creating / running the container and was able to successfully run the test case from the quick start
Code:
[user@cesm2.2 ~]$ create_newcase --case test1 --compset F2000climo --res f19_g17 --run-unsupported
Compset longname is 2000_CAM60_CLM50%SP_CICE%PRES_DOCN%DOM_MOSART_CISM2%NOEVOLVE_SWAV
Compset specification file is /opt/ncar/cesm2/components/cam//cime_config/config_compsets.xml
Automatically adding SIAC to compset
Automatically adding SESP to compset
Compset forcing is 1972-2004
ATM component is CAM cam6 physics:
LND component is clm5.0:Satellite phenology:
ICE component is Sea ICE (cice) model version 5 :prescribed cice
OCN component is DOCN   prescribed ocean mode
ROF component is MOSART: MOdel for Scale Adaptive River Transport
GLC component is cism2 (default, higher-order, can run in parallel):cism ice evolution turned off (this is the standard configuration unless you're explicitly interested in ice evolution):
WAV component is Stub wave component
IAC component is Stub iac component
ESP component is Stub external system processing (ESP) component
Pes     specification file is /opt/ncar/cesm2/components/cam//cime_config/config_pes.xml
Compset specific settings: name is SSTICE_DATA_FILENAME and value is $DIN_LOC_ROOT/atm/cam/sst/sst_HadOIBl_bc_1x1_2000climo_c180511.nc
Could not find machine match for 'de9bc9acbce1' or 'de9bc9acbce1'
Machine is container
Pes setting: machine match is container
Pes setting: grid          is a%1.9x2.5_l%1.9x2.5_oi%gx1v7_r%r05_g%gland4_w%null_z%null_m%gx1v7
Pes setting: compset       is 2000_CAM60_CLM50%SP_CICE%PRES_DOCN%DOM_MOSART_CISM2%NOEVOLVE_SWAV_SIAC_SESP
Pes setting: tasks       is {'NTASKS_ATM': -1, 'NTASKS_LND': -1, 'NTASKS_ROF': -1, 'NTASKS_ICE': -1, 'NTASKS_OCN': -1, 'NTASKS_GLC': -1, 'NTASKS_WAV': -1, 'NTASKS_CPL': -1}
Pes setting: threads     is {'NTHRDS_ATM': 1, 'NTHRDS_LND': 1, 'NTHRDS_ROF': 1, 'NTHRDS_ICE': 1, 'NTHRDS_OCN': 1, 'NTHRDS_GLC': 1, 'NTHRDS_WAV': 1, 'NTHRDS_CPL': 1}
Pes setting: rootpe      is {'ROOTPE_ATM': 0, 'ROOTPE_LND': 0, 'ROOTPE_ROF': 0, 'ROOTPE_ICE': 0, 'ROOTPE_OCN': 0, 'ROOTPE_GLC': 0, 'ROOTPE_WAV': 0, 'ROOTPE_CPL': 0}
Pes setting: pstrid      is {}
Pes other settings: {}
Compset is: 2000_CAM60_CLM50%SP_CICE%PRES_DOCN%DOM_MOSART_CISM2%NOEVOLVE_SWAV_SIAC_SESP
Grid is: a%1.9x2.5_l%1.9x2.5_oi%gx1v7_r%r05_g%gland4_w%null_z%null_m%gx1v7
Components in compset are: ['cam', 'clm', 'cice', 'docn', 'mosart', 'cism', 'swav', 'siac', 'sesp']
No project info available
No charge_account info available, using value from PROJECT
cesm model version found: cesm2.2.0
Batch_system_type is none
Creating Case directory /home/user/test1
[user@cesm2.2 ~]$ ls
test1
Thanks for the brief explanation!!! now i am able to run the test cases.
 

Yonas

Yonas
New Member
Hi Ryanj, i am trying to run a single point "PTS_MODE" simulation using the following command as stated in the manual but i am getting this error: create_newcase: error: unrecognized arguments: -pts_lat 40.0 -pts_lon -105

This is the code i used:
create_newcase -case testPTS_MODE -res f19_g16 -compset I1850CRUCLM45BGC -pts_lat 40.0 -pts_lon -105

It means the create_newcase doesn't accept the pts_lat and pts_lon arguments, any help on this issue & here is the guide i used to create the single point "PTS_MODE" simulation. 1.6.2. Running a single point using global data - PTS_MODE — ctsm CTSM master documentation

Thanks,
 

dobbins

Brian Dobbins
CSEG and Liaisons
Staff member
It means the create_newcase doesn't accept the pts_lat and pts_lon arguments, any help on this issue & here is the guide i used to create the single point "PTS_MODE" simulation. 1.6.2. Running a single point using global data - PTS_MODE — ctsm CTSM master documentation

Thanks,

Hi Yonas,

I think there's two issues here - the first is an error in our documentation (oops!). The --pts_lat / --pts_lon commands were removed from the create_newcase script a while ago, so the way to do this now is via the xmlchange utility after creating a case, eg:

./xmlchange PTS_MODE=TRUE,PTS_LAT=40.0,PTS_LON=105

The second issue is I think you're using a non-existent compset? I'm not terribly familiar with the land compsets, but grepping for that one it only shows up in a ChangeLog and in a 'sed' script showing a switch to 'I1850Clm50BgcCropCru'. If that's not what you want, I can try to find someone who can provide more information, or the land-model section of the forum might be able to provide answers.

That said, if I do those two commands, it creates and builds for me (though see below for why we can't run yet):

create_newcase --case testPTS_MODE --res f19_g16 --compset I1850Clm50BgcCropCru --run-unsupported
cd testPTS_MODE
./xmlchange PTS_MODE=TRUE,PTS_LAT=40.0,PTS_LON=105
./case.setup
./case.build

Note that I also switched to double-hyphens, which is the new style in CESM2, and added the '--run-unsupported' flag, since this compset/resolution isn't scientifically supported in CESM 2.2.

At this point, though, you're going to hit a problem running it - specifically, when you try to download the input data, it doesn't exist on the FTP servers. (It does exist on our Cheyenne system, if you have access to that.). I'll check with the land model team about that in the morning. Finally, just be aware that the forcing files for these runs are huge -- the tested compsets listed in the first post show that the Vancouver single-point case only needed 7-11GB, but I think that might be a special case, with others needing to download the hundreds of gigs of input data. The good news is any additional runs of this kind will already have that (assuming you're mounting a directory into the container!), but it's definitely a big hurdle to get past for initial test runs. I'll double check that, too, and get back to you.

Hope that helps, and let me know if you run into more issues or have other questions -- we'll fix the documentation soon, too!

Cheers,
- Brian
 

Yonas

Yonas
New Member
Hi Yonas,

I think there's two issues here - the first is an error in our documentation (oops!). The --pts_lat / --pts_lon commands were removed from the create_newcase script a while ago, so the way to do this now is via the xmlchange utility after creating a case, eg:



The second issue is I think you're using a non-existent compset? I'm not terribly familiar with the land compsets, but grepping for that one it only shows up in a ChangeLog and in a 'sed' script showing a switch to 'I1850Clm50BgcCropCru'. If that's not what you want, I can try to find someone who can provide more information, or the land-model section of the forum might be able to provide answers.

That said, if I do those two commands, it creates and builds for me (though see below for why we can't run yet):



Note that I also switched to double-hyphens, which is the new style in CESM2, and added the '--run-unsupported' flag, since this compset/resolution isn't scientifically supported in CESM 2.2.

At this point, though, you're going to hit a problem running it - specifically, when you try to download the input data, it doesn't exist on the FTP servers. (It does exist on our Cheyenne system, if you have access to that.). I'll check with the land model team about that in the morning. Finally, just be aware that the forcing files for these runs are huge -- the tested compsets listed in the first post show that the Vancouver single-point case only needed 7-11GB, but I think that might be a special case, with others needing to download the hundreds of gigs of input data. The good news is any additional runs of this kind will already have that (assuming you're mounting a directory into the container!), but it's definitely a big hurdle to get past for initial test runs. I'll double check that, too, and get back to you.

Hope that helps, and let me know if you run into more issues or have other questions -- we'll fix the documentation soon, too!

Cheers,
- Brian
Hi Brian, thanks for your explanation, Now I can build the PTS_MODE following your instructions & I used the Vancouver single-point case since the previous composite could be depreciated as you mentioned earlier. This is the code I used to build the case & it works fine...

create_newcase --case testPTS_MODE --res f19_g16 --compset I1PtClm50SpRsGs --run-unsupported
cd testPTS_MODE
./xmlchange PTS_MODE=TRUE,PTS_LAT=40.0,PTS_LON=105
./case.setup
./case.build
./case.submit

but after case.submit , I am facing a problem in downloading all the input data, I attached the error message so that you can see what went wrong with my run.
After facing this issue, i used the "./check_input_data" command to identify the missing data & i downloaded most of them manually to the inputdata folder but there is missing data related to the "datm" file since I am getting this error message:

Model datm missing file domain1 = 'null/null'
Model datm missing file file1 = 'null/null'

Thanks in advance for your help!!!
 

Attachments

  • ERROR.txt
    5.6 KB · Views: 3

dobbins

Brian Dobbins
CSEG and Liaisons
Staff member
Hi Brian, thanks for your explanation, Now I can build the PTS_MODE following your instructions & I used the Vancouver single-point case since the previous composite could be depreciated as you mentioned earlier. This is the code I used to build the case & it works fine...

create_newcase --case testPTS_MODE --res f19_g16 --compset I1PtClm50SpRsGs --run-unsupported
cd testPTS_MODE
./xmlchange PTS_MODE=TRUE,PTS_LAT=40.0,PTS_LON=105
./case.setup
./case.build
./case.submit

but after case.submit , I am facing a problem in downloading all the input data, I attached the error message so that you can see what went wrong with my run.
After facing this issue, i used the "./check_input_data" command to identify the missing data & i downloaded most of them manually to the inputdata folder but there is missing data related to the "datm" file since I am getting this error message:

Model datm missing file domain1 = 'null/null'
Model datm missing file file1 = 'null/null'

Thanks in advance for your help!!!

Hi Yonas,

I think the best thing I can do is recommend posting in the CTSM/CLM forum section here:
CTSM/CLM Forum Page

The reason for this is that this behavior you're encountering is not unique to the container, as the same thing happens on our Cheyenne supercomputer - so it's a land-model issue, not a technical one. That said, I think the primary issue is that CESM doesn't (yet) automatically subset data to provide inputs for arbitrary points. So running something like the 'I1PtClm50SpRsGs' compset with the 1x1_vancouverCAN resolution works, because those mappings are set up, but running arbitrary points like 40.0 / 105 has no default mapping, and you need to generate, then specify the inputs.

In other words, if you just want to try running a model, try this:
create_newcase --case testPTS_Vancouver --res 1x1_vancouverCAN --compset I1PtClm50SpRsGs --run-unsupported
cd testPTS_Vancouver
./xmlchange NTASKS=1
./case.setup
./case.build
./case.submit

(Note that I'm explicitly setting the number of tasks to 1 -- this is needed in the container due to a bug, set to be fixed in the upcoming build. It only applies to single-point cases!)

And that should work fine, both on the container and on Cheyenne. But when you start switching out the pre-built locations for arbitrary ones, it gets a bit trickier. I'm not an expert on the land model modes, so posting there will likely get you far better support than I can offer. If you do run into additional technical, things, though, let me know and I'm happy to help!
 

Yonas

Yonas
New Member
Hi Yonas,

I think the best thing I can do is recommend posting in the CTSM/CLM forum section here:
CTSM/CLM Forum Page

The reason for this is that this behavior you're encountering is not unique to the container, as the same thing happens on our Cheyenne supercomputer - so it's a land-model issue, not a technical one. That said, I think the primary issue is that CESM doesn't (yet) automatically subset data to provide inputs for arbitrary points. So running something like the 'I1PtClm50SpRsGs' compset with the 1x1_vancouverCAN resolution works, because those mappings are set up, but running arbitrary points like 40.0 / 105 has no default mapping, and you need to generate, then specify the inputs.

In other words, if you just want to try running a model, try this:


(Note that I'm explicitly setting the number of tasks to 1 -- this is needed in the container due to a bug, set to be fixed in the upcoming build. It only applies to single-point cases!)

And that should work fine, both on the container and on Cheyenne. But when you start switching out the pre-built locations for arbitrary ones, it gets a bit trickier. I'm not an expert on the land model modes, so posting there will likely get you far better support than I can offer. If you do run into additional technical, things, though, let me know and I'm happy to help!
Thanks a lot Brian, i will post the previous issue on the CTSM/CLM forum as you proposed. Regarding the above example, 'I1PtClm50SpRsGs' compset with the 1x1_vancouverCAN, i followed your steps and it works fine except for the ./case.submit step. The error is related to the 'Archiving restarts for date' you can see the error message i got. I think a single point run doesn't support restarts but i am not sure about this.

Additionally, I need your advice on using the CESM container you built for my Ph.D. project, which is proposed to use the CLM model. Initially, I was trying to install the CESM or CTSM on my laptop and I was facing some problems during the installation related to the libraries. Suddenly, I came across the CESM container, and I tried to experiment on it and it works fine, So can I fully depend on the container for my project without installing the CESM/CTSM or there are other options?

Thanks in Advance!!!
 

Attachments

  • restart.txt
    5.6 KB · Views: 2

dobbins

Brian Dobbins
CSEG and Liaisons
Staff member
Thanks a lot Brian, i will post the previous issue on the CTSM/CLM forum as you proposed. Regarding the above example, 'I1PtClm50SpRsGs' compset with the 1x1_vancouverCAN, i followed your steps and it works fine except for the ./case.submit step. The error is related to the 'Archiving restarts for date' you can see the error message i got. I think a single point run doesn't support restarts but i am not sure about this.

I think that board might have more to say on that warning too - note that it is a warning, not an error. The restarts are indeed placed in the ~/archive directory, and this error also pops up in other scenarios, as seen in this post from a short while ago (without a response yet, though):

https://bb.cgd.ucar.edu/cesm/thread...lure-err-ncdump-locfnh-no-such-variable.5798/

Additionally, I need your advice on using the CESM container you built for my Ph.D. project, which is proposed to use the CLM model. Initially, I was trying to install the CESM or CTSM on my laptop and I was facing some problems during the installation related to the libraries. Suddenly, I came across the CESM container, and I tried to experiment on it and it works fine, So can I fully depend on the container for my project without installing the CESM/CTSM or there are other options?

This I can comment on! Short version, yes - this container passes what is called the 'Ensemble Consistency Test' (ECT) used in validating a port of CESM. While this version is being called a 'technical preview', we will have an official release in the future, too, and expect those to track future versions of CESM. The first difference between that first official release and this one is that some of the (very minor) changes we made to get it working easily in the container will be folded back into the CESM source code, which means at that point, you can easily do a 'git clone' of the CESM source, even development versions, and it should work on the container. That might be helpful for your PhD if you're modifying code. (Note that you can do this now, too, as you can do 'sudo /bin/bash' in the container, switching to root, and editing the source code directly... but it gets a lot easier when you can just check out code as yourself!).

The second thing is we're awaiting some word on what we can do with regards to compilers - currently, the image uses GNU Fortran, but the Intel compilers are now available for free, and typically much faster. For the land model this might be 20%-50% faster. If you're doing runs of any length, that might be something of interest. Right now, it's just a legal question of whether we can embed them, or you need to download them yourself, but I should know more in early January.

So, yes, you can depend on it for your thesis, though depending on the complexity of your runs (eg, if it's all single-point runs, or if you're doing global runs, or how long they're going for, etc.), maybe we should talk at some point. If you're doing short, single-point runs, your laptop is probably fine. If you're doing hundreds or thousands of runs, or long runs, maybe access to Cheyenne (the NCAR supercomputer) or our upcoming cloud (eg, AWS) version of CESM might help too.

Hope that helps, and let me know if you have more questions!
- Brian
 

Yonas

Yonas
New Member
I think that board might have more to say on that warning too - note that it is a warning, not an error. The restarts are indeed placed in the ~/archive directory, and this error also pops up in other scenarios, as seen in this post from a short while ago (without a response yet, though):

https://bb.cgd.ucar.edu/cesm/thread...lure-err-ncdump-locfnh-no-such-variable.5798/



This I can comment on! Short version, yes - this container passes what is called the 'Ensemble Consistency Test' (ECT) used in validating a port of CESM. While this version is being called a 'technical preview', we will have an official release in the future, too, and expect those to track future versions of CESM. The first difference between that first official release and this one is that some of the (very minor) changes we made to get it working easily in the container will be folded back into the CESM source code, which means at that point, you can easily do a 'git clone' of the CESM source, even development versions, and it should work on the container. That might be helpful for your PhD if you're modifying code. (Note that you can do this now, too, as you can do 'sudo /bin/bash' in the container, switching to root, and editing the source code directly... but it gets a lot easier when you can just check out code as yourself!).

The second thing is we're awaiting some word on what we can do with regards to compilers - currently, the image uses GNU Fortran, but the Intel compilers are now available for free, and typically much faster. For the land model this might be 20%-50% faster. If you're doing runs of any length, that might be something of interest. Right now, it's just a legal question of whether we can embed them, or you need to download them yourself, but I should know more in early January.

So, yes, you can depend on it for your thesis, though depending on the complexity of your runs (eg, if it's all single-point runs, or if you're doing global runs, or how long they're going for, etc.), maybe we should talk at some point. If you're doing short, single-point runs, your laptop is probably fine. If you're doing hundreds or thousands of runs, or long runs, maybe access to Cheyenne (the NCAR supercomputer) or our upcoming cloud (eg, AWS) version of CESM might help too.

Hope that helps, and let me know if you have more questions!
- Brian
Hi Brian, thanks for your willingness to answer my questions and its really helpful. The 'I1PtClm50SpRsGs' compset with the 1x1_vancouverCAN is working, it was just a warning, not an error as you said. i just made a 30 year run with daily time steps.

About the container i am glad its suits for my project, i will also follow the future releases. About my project, the plan is to make regional runs by CLM to study some land surface properties at a country level in Ethiopia at different spatial resolutions (say 0.2, 0.1, 0.5 degrees). For the historical period a 30 year run will be enough using high resolution ERA5 atmospheric forcing dataset and for future period 80 year run is proposed. I am not sure how many years i can simulate per day with the laptop. Can you give me some advise on how i can get access to the computing facilities you mentioned free of charge?

Thanks!
 
Top