-
Notifications
You must be signed in to change notification settings - Fork 87
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Make thompson_mynn_lam3km ccpp suite available #644
Make thompson_mynn_lam3km ccpp suite available #644
Conversation
Hi @chan-hoo, @JiliDong-NOAA, should this new aerosol dataset used by Thompson replace the climatology that is currently being provided in chgres_cube? The workflow flips the "thomp_mp_climo_file" flag in the chgres_cube namelist to true if Thompson is chosen and if the ICs and LBCs don't come from the RAP or HRRR, which contain aerosol data (see the ex-scripts for ICs/LBCs for this logic). So chgres_cube creates the necessary climo fields in that case. How do these changes work with that climatology? |
@JiliDong-NOAA, Can you answer to @JeffBeck-NOAA's comment? And, should I move 'FV3_GFS_v15_thompson_mynn_lam3km' from Ln 182 to between Lns 190-191 in exregional_make_ics.sh (and same change in exregional_make_lbcs.sh)? |
@chan-hoo, I believe you will want to move the SDF from line 182 to ~190-191 in the ex-script IC/LBC creation files, since the main goal there is to correctly handle the aerosols for Thompson. @LarissaReames-NOAA, we should eventually overhaul/rename the varmap tables in UFS_UTILS to no longer be based on "GFS" or "GSL" physics, since people are now running more complicated combinations. @chan-hoo, after you make this change, you'll want to confirm that your gfs_data.nc file, created by chgres_cube, contains the correct aerosol (and other) fields. |
@JeffBeck-NOAA @chan-hoo the MERRA2 aero clim dataset is for the independent aerosol module/scheme and not particular for Thompson scheme. I don't know if we can replace "thomp_mp_climo_file" with the MERRA2 dataset at this point. I will double check with our aerosol expert Anning on this and get back to you. |
Thanks, @JiliDong-NOAA. So Thompson MP doesn't use this new MERRA2 dataset at all? If not, what scheme uses it? |
@JeffBeck-NOAA the MERRA2 aerosol dataset is for aerosol scheme which is related to but different from microphysics scheme. MERRA2 aerosol can be combined with Thompson MP or GFDL MP. It is also confirmed that we still need to use Thompson_MP_MONTHLY_CLIMO.nc to initialized water/ice friendly aerosol for Thompson MP scheme for now. |
Thanks, @JiliDong-NOAA! @chan-hoo, did your test succeed and produce the necessary tracers/aerosols (ice- and water-friendly) when you moved the SDF in the chgres_cube ex-scripts? If you're using non-RAP/HRRR data as ICs/LBCs, you should have gotten the climo aerosol distribution. |
@JeffBeck-NOAA, with HRRR/RAP, the test run failed with: "fms_io(restore_state_all): unable to find mandatory variable tref in restart file sfc_data.nc". Should I move the SDF line back to Line 182? or are there any other solutions? with FV3GFS/FV3GFS, the test run was complete successfully. |
@JeffBeck-NOAA, thompson_mynn_lam3km does not run with RAP/HRRR data at this moment. What about putting an error message when it is selected with RAP/HRRR in setup.sh? |
@chan-hoo, did you try moving the suite back to line 182? When you got the error about tref, what does your |
@JeffBeck-NOAA, yes, I did. But I got the same error. In input.nml, "nstf_name = 2, 0, 0, 0, 0". |
xkzminv: 0.3 | ||
xkzm_m: 1.0 | ||
xkzm_h: 1.0 | ||
|
||
FV3_GFS_v16: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@chan-hoo, This SDF won't work with RAP/HRRR-based ICs unless "nstf_name" is 2, 1, 0, 0, 0. Does it need to be 2, 0, 0, 0, 0, or can we change it?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I believe that switching the second flag in nstf_name from 0 to 1 means that the NSST data from the surface data file will not be read into the model. If this is not desired/needed, I don't see an issue with changing nstf_name to 2,1,0,0,0
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@BenjaminBlake-NOAA, do you have an answer to Jeff's comment? I'll test [2,1,0,0,0] with both FV3GFS and RAP/HRRR.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@chan-hoo I think we added comments at the same time :)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, @BenjaminBlake-NOAA, @chan-hoo, it looks like we need a switch in the workflow to change the values of nstf_name based on the external model data being used for ICs. Since we don't have that at the moment, the easiest solution is to set all suites to 2,1,0,0,0, in which case, I believe FV3 will spin up its own NSST values.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, @BenjaminBlake-NOAA. I'll test them.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@JeffBeck-NOAA Yes I think that's right about FV3 spinning up its own NSST values
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@JeffBeck-NOAA, @BenjaminBlake-NOAA, both FV3GFS and RAP/HRRR works well with 2,1,0,0,0.
@@ -593,6 +593,25 @@ fi | |||
# | |||
#----------------------------------------------------------------------- | |||
# | |||
# Copy MERRA2 aerosol climatology data. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@chan-hoo @JeffBeck-NOAA Question: Do these directories need to always be copied or only when using the thompson_mynn_lam3km suite? If the latter, can we wrap them in an if-statement that checks for the suite (i.e. in a new PR, or I can throw it in my current PR)? Just want to reduce the time it takes to generate a new experiment... Thanks.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@gsketefian, Jili answered this question above: "the MERRA2 aero clim dataset is for the independent aerosol module/scheme and not particular for Thompson scheme."
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@chan-hoo, is the MERRA2 aero clim dataset only used when running RRFS-CMAQ? If so, then, we may want something like DO_CMAQ="true" in the config.sh file to avoid copying these files unless it's being used.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@JeffBeck-NOAA, No. Since RRFS-CMAQ is coupled with GFS v15, CCPP is not available. MERR2 is used in SRW App (FV3-LAM).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@chan-hoo, right, so CMAQ isn't in the picture. So what component of the SRW App uses the MERRA2 dataset?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@JeffBeck-NOAA, if I understand correctly, it is used in thompson_mynn_lam3km now but it can be selected by other CCPP suites. @JiliDong-NOAA, is this correct? Can we turn it on/off in the fv3 namelist?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@gsketefian @JeffBeck-NOAA , I'll put a flag to turn on/off MERRA2 like 'DO_MERRA' into the regional workflow soon.
DESCRIPTION OF CHANGES:
TESTS CONDUCTED:
and a new WE2E test on WCOSS dell and cray, and Hera:
DEPENDENCIES:
ISSUE:
CONTRIBUTORS:
@BenjaminBlake-NOAA , @JiliDong-NOAA