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

Output of SMB in CLM4.5 F-compset runs

wadewei

Wade Wei
New Member
Hi everyone,

I am trying to have CLM4.5 (within CESM1.2.2) output surface mass balance in the Himalayas. However, it seems that CLM in default will only calculate SMB for runs coupled with CISM. One concern is that f05 resolution seems to be unsupported out of box with CESM runs with CISM such as F_1850_CN_GLC_CISM1, and there seems to only have f09 and f19 glcmaskdata in inputdata/glc/cism/griddata/.

I tried to xmlchange GLC_NEC from 0 to 10 in my existing F1850 run with f05 resolution, but it would give me an error stating glacial mask is needed. I wonder if if would work if I create an f05 glcmask with ones in the Himalayas? Additionally, I assume GLC_SMB would also have to be changed to true?

Or is there an alternative way to get what I need?

Thank you,
Wade
 

sacks

Bill Sacks
CSEG and Liaisons
Staff member
Yes, there is limited support for outputting SMB in CESM1.2. I forget all of the limitations, but remember that there were indeed limitations like this.

CESM1.2.2 is no longer supported, but the CESM2 series (which is supported) has more flexible support for what you want to do here. (See CESM2 Release Series Information for some information on this release series.) Is it feasible for you to update to CESM2 for your work?
 

wadewei

Wade Wei
New Member
Yes, there is limited support for outputting SMB in CESM1.2. I forget all of the limitations, but remember that there were indeed limitations like this.

CESM1.2.2 is no longer supported, but the CESM2 series (which is supported) has more flexible support for what you want to do here. (See CESM2 Release Series Information for some information on this release series.) Is it feasible for you to update to CESM2 for your work?
I will look into CESM2 for upgrade possibility. Could you point me to some sort of transition guide if there is any, especially those related to paleoclimate simulations?

However, one strong motivation for us to remain with CESM1.2.2 and CAM4 is the computation cost, which almost triples with CAM5. We may have to stick with 1.2.2 for a while. Do you think it's possible to output SMB without making major changes to the model?
 

sacks

Bill Sacks
CSEG and Liaisons
Staff member
I'm not aware of anything that is really a transition guide exactly. You can find a description of major changes here CESM2 Release Series Information , and User's Guides for the components linked from Community Earth System Model - CESM2 . Especially see the CIME User's Guide for general user interface / scripts documentation. I'd suggest asking paleoclimate-specific questions in Paleoclimate . I'll also mention that CESM1.2 is still supported in a limited capacity for deep time paleoclimate simulations; if that's what you're running, then again the paleoclimate forums would be a good place to try to get some support.

My understanding is that CAM4 is still available in CESM2. There may not be many out-of-the-box compsets, but you can still select CAM4 via a user-defined compset. From skimming <4. Atmospheric configurations (compsets) — camdoc documentation>, though, it sounds like CAM4 in CESM2 is not exactly the same as CAM4 in CESM1.2, so I'm not sure if it's considered scientifically recommendable to use it. You could ask that question on the paleoclimate or CAM forums.

If you do stick with CESM1.2, we aren't going to be able to provide much support, unfortunately: A lot has changed in the model since then, and we have limited support resources, so we aren't able to provide much support for use of these old versions. My memory isn't very good about exactly what was possible in these older versions, but I think it may be hard to force the model to produce SMB output unless you include CISM. So I think your best bet is to compare a f09 case with this f05 case and see what files you need to provide in order to run CISM at f05 resolution. For example, you will need mapping files between f05 and the CISM grid, which you can create with the mapping file generation utilities in that code base. It may be challenging to provide all the files you need, and unfortunately we aren't able to provide support with that process for CESM1.2. This is all significantly easier in CESM2, where you no longer need CISM to produce SMB.

Regardless of which model version you use, though, I should caution you that past attempts to get reasonable SMB over the Himalayas have been very unsuccessful. Specifically, it has been very hard to get the model to produce positive SMB in these high mountain regions. There is some promising recent work within CESM2 development code, but that relies on (a) running variable-resolution cases with high resolution focused over this region, (b) updating the glacier and topography datasets to be more accurate in this region, and (c) some changes to the code. Without all of these changes, I think you may spend a lot of time trying to get this working only to find that you can't use it for the science you want to do.

One other possible issue with trying to run over this region is the number of glacier elevation classes. The default of 10 doesn't do a good job of capturing the elevation gradients in this high mountain region. The CESM1.2 code base may (or may not) include partial support for 36 elevation classes, which would be better (though more expensive to run); you would need to generate your own surface dataset using CLM's mksurfdata_map tool (which I think may allow you to specify 36 elevation classes), would need to change GLC_NEC to 36, and may need to make some other changes.

If you decide to pursue this despite these caveats, then I'm afraid I won't be able to help further with CESM1.2, since what I've said above is about the limit of what I can remember about this model version.
 
Top