8.2.2-2.10 Fix for variable density snow accumulation with RUC LSM #104
Conversation
1. Fix the bug that crashes MPAS when frac_seaice=false. Use 'if(config_frac_seaice) then' around arrays with *_sea that are allocated only when config_frac_seaice=true. This bug is reported to Michael Duda. 2. Add flqc and flhc computation for fractional ice. These variables are used in RUC ice module and are needed for reproducibility over sea ice.
1. Do not deallocate flqc_sea and flhc_sea - they are used later in the seaice driver.
xice=1. It will reduce computations over solid sea ice (Arctic).
CPM is not RUC-specific variable, moved out of sf_ruc IF.
…into seaice_fix
…into seaice_fix
background albedo when water is frozen.
lines for easy merge with NCAR code.
In the LSM driver frpcpn = false. If it is false, frozen precipitation in RUC and Noah LSMs is computed using temperature threshold and ignoring the information from the microphysics scheme. A new parameter frpcpn_ruc is introduced for use with RUC LSM, and it is set to true. This change will use the microphysics information correctly and will fix the problem with the run-total accumulation of snow depth.
|
I will redo this PR after PR#97 is merged. This PR should have a change only src/core_atmosphere/physics/mpas_atmphys_driver_lsm.F. |
tanyasmirnova
changed the title
jderrico-noaa
requested review from
joeolson42,
haiqinli,
mdtoyNOAA,
AndersJensen-NOAA and
clark-evans
There was a problem hiding this comment.
We need to identify why the answers do not change at 12 minutes but do by 60 minutes into the hrrrv5 simulations. @joeolson42 suggests it could be that the accumulation differences are too small to trigger the surface layer driver that would change the results. If we're comfortable with that, then I'm OK merging. Otherwise, further investigation is warranted.
I have explained it in the Chat last week. The time step in our reg. tests is 12 min. At initial time precipitation is zero, therefore there is no difference after one time step. |
Thank you for clarifying, @tanyasmirnova - I did not fully understand that from the chat last time. The model integrates forward for 12 min. Everything should be the same except for the resulting snow fields. Please correct me if I'm wrong, but I naively presume those snow fields are updated after physics tendencies are calculated, such that any differences in the resulting snow fields will only begin to impact the model solution after this first time step? |
|
Snow accumulation depth is calculated in the RUC LSM. During the first time step there is no precipitation (liquid or frozen) when RUC LSM is called, it will be computed later in the microphysics. Thus, the change in this PR will affect the results starting only from time step #2. This is why we see identical results between the 12-min outputs. |
|
@tanyasmirnova you should merge top of develop into this branch since there are merge conflicts because of the version differences in the three files noted below. |
tanyasmirnova
force-pushed
the
RUC_snowfallac_fix
branch
from
fce68a9 to
08f6755
Compare
merge gsl/develop into RUC_snowfallac_fix
The title above should be a 1 line short summary of the pull request (i.e. what the project the PR represents is intended to do).
Enter a description of this PR. This should include why this PR was created, and what it does.
Testing and relations to other Pull Requests should be added as subsequent comments.
See the below examples for more information.
MPAS-Dev/MPAS#930
MPAS-Dev/MPAS#931
Information on running mandatory regression tests on Jet can be found here and the results pasted below.
Regression test is ran against version 8.2.2-2.8. The results differ only in the hrrrv5 suite as expected.
The compare file is in /mnt/lfs5/BMC/wrfruc/smirnova/MPAS_regression_tests/mpas_testcase/utils/compare_run_testcases_v8.2.2-2.10-intdebug
regression test case results