dynamic landunits: handle methane with changing lake area

[billsacks]

Bill Sacks < sacks@ucar.edu > - 2016-02-16 13:38:10 -0700 Bugzilla Id: 2283 Bugzilla CC: erik@ucar.edu, oleson@ucar.edu, rfisher@ucar.edu,

The methane code (ch4Mod.F90) implicitly assumes that finundated is 1 for lakes. Thus, if lake area increases, the right thing to do is probably to immediately inundate any land it took over, with code similar to what happens right now when inundated area increases.

Decreases in lake area should be handled appropriately now, but increases in lake area are not. Currently, there is no way for lakes to increase in area, but this will need to be revisited once they can.

[billsacks]

When trying to run a dynamic lakes BGC case (specifically, ERS_Lm25.1x1_smallvilleIA.IHistClm50BgcCropQianRsGs.cheyenne_gnu.clm-smallville_dynlakes_monthly from 3ab74de5), I am getting a gridcell methane conservation issue. I haven't dug into it, but I'm thinking it may be related to this issue.

[billsacks]

Now that dynamic lakes are nearly on master, I'm changing this from an enhancement to a bug.

[billsacks]

Discussion in ctsm-software meeting: For now, in order to allow ch4 always on with BGC: Let's turn off the gridcell methane conservation check when dynamic lakes are active, maybe just in the first time step of each year.

Then we'll come back to this later and fix it properly. But since the conservation errors are probably quite small, this isn't high priority.

[slevis-lmwg]

This issue describes a ch4 balance error with transient lakes. I see a ch4 balance error in #1868 but:

• I have yet to confirm that my problem is due to transient lakes and
• My balance error appears in hybrid/startup cases (e.g. 1901, 1920, 1921) though not in 1850 where the whole simulation started and not in restart/branch cases.

The way the problem manifests in #1868 and if it's the same as #43, we could comment out the gridcell balance check (as suggested above) but for the first timestep of the run.

[billsacks]

@slevisconsulting pointed out that we have this current workaround in place to bypass the gridcell-level methane checks on the first time step of each year when using dynamic lakes:

https://github.com/ESCOMP/CTSM/blob/616905bbbac55f506fccc5aef8c45bbbdbba94a2/src/biogeochem/ch4Mod.F90#L2330-L2332

together with some code that bypasses the ch4 balance checks if this flag is true.

As @slevisconsulting pointed out in https://github.com/ESCOMP/CTSM/discussions/1868#discussioncomment-3901132 it's likely that there is a problem in time step 0 of a hybrid case that may need a similar workaround to the one above.