Use USTAR read from GEOS instead of calculating from U10M and V10M #279
Conversation
yantosca
added
topic: Input Data
There was a problem hiding this comment.
Thanks @yuanjianz. We can merge this pending updating the CHANGELOG.md file.
src/Extensions/hcox_dustdead_mod.F - Commented out unused codes for shifting to use USTAR read from meteorology CHANGELOG.md - Added a description: Use USTAR from meteorology instead of calculating from reference 10m wind in DustDead extension
|
Also, this PR will need to go into GC 14.5.0 / HEMCO 3.10.0, as this will change the fullchem benchmark simulation results. I've changed the target branch accordingly. |
There was a problem hiding this comment.
Thanks @yuanjianz. I have a few minor comments, mostly about removing obsolete commented-out code.
|
@lizziel: If we remove the code in the dust extension to compute USTAR (friction velocity), would this prove to be a problem for other models like CESM? I presume that is one of the meteorological variables available. |
|
@yantosca: Removing the computation of USTAR in the dust extension is fine since we already have it available for use in drydep. |
There was a problem hiding this comment.
Thanks @yuanjianz. Good to merge. I'll start testing this now.
This merge brings in the updates for HEMCO PR #279 (Use USTAR read from GEOS instead of calculating from U10M and V10M, by @yuanjianz) atop the HEMCO 3.9.1 release. This will be merged into the HEMCO 3.10.0 development stream after passing integration tests. Signed-off-by: Bob Yantosca <[email protected]>
|
@yuanjianz @ltmurray: Would you know what the scaling factor would be for the GCAP2 meteorology? |
|
I don't know what the scaling factor would be for GCAP2, which is presently tuned given the existing parameterization. USTAR has been archived for GCAP2 though. |
|
@yuanjianz @msulprizio @ltmurray @lizziel: We would need a companion PR to go along with this to update the setting of the The current settings in #----------------------------------------------------------------
# Horizontal resolution-dependent settings
#-----------------------------------------------------------------
if [[ ${met} = "ModelE2.1" ]]; then
if [[ "$runid" == "E213f10aF40oQ40nudge" ]]; then
if [[ "$grid_res" == "4x5" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.00474046'\n"
elif [[ "$grid_res" == "2x25" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.00243979'\n"
elif [[ "$grid_res" == "05x0625" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.00276896'\n"
elif [[ "$grid_res" == "025x03125" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.00254319'\n"
else
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='-999.0e0'\n"
fi
else
if [[ "$grid_res" == "4x5" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.03564873'\n"
elif [[ "$grid_res" == "2x25" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.01050036'\n"
elif [[ "$grid_res" == "05x0625" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.01340854'\n"
elif [[ "$grid_res" == "025x03125" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='0.01066495'\n"
else
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='-999.0e0'\n"
fi
fi
else
RUNDIR_VARS+="RUNDIR_GISS_RES='not_used'\n"
# Use GEOS-FP values as placeholders for GEOS-IT until parameters derived
if [[ "x${sim_name}" == "xfullchem" || "x${sim_name}" == "xaerosol" ]]; then
if [[ "x${met}" == "xgeosfp" && "x${grid_res}" == "x4x5" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='8.3286e-4'\n"
elif [[ "x${met}" == "xgeosfp" && "x${grid_res}" == "x2x25" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='5.0416e-4'\n"
elif [[ "x${met}" == "xmerra2" && "x${grid_res}" == "x4x5" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='7.8533e-4'\n"
elif [[ "x${met}" == "xmerra2" && "x${grid_res}" == "x2x25" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='4.7586e-4'\n"
elif [[ "x${met}" == "xgeosit" && "x${grid_res}" == "x4x5" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='8.3286e-4'\n"
elif [[ "x${met}" == "xgeosit" && "x${grid_res}" == "x2x25" ]]; then
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='5.0416e-4'\n"
else
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='-999.0e0'\n"
fi
else
RUNDIR_VARS+="RUNDIR_DUSTDEAD_TF='-999.0e0'\n"
fi
fi |
|
@yuanjianz @yidant: Would either of you be able to compute the dust scale factors (see above post)? |
Hi @yantosca, sorry for the late response. I will try to work on this with Yidan. @yidant, do we have GCAP2 archive at WashU? |
|
Thanks @yidant & @yuanjianz. When you can get to this. Not immediately urgent. |
|
Hi @yantosca, I haven't got the chance to touch base on GCAP2 yet. But I have already calculated dust scale factors for 2x2.5 and 4x5 for MERRA-2, GEOS-FP and GEOS-IT. See geoschem/geos-chem#2512. Tag @ltmurray @yidant for help of running HEMCO standalone with GCAP2 meteorology. |
|
Hi @yantosca @yuanjianz and @ltmurray, we discussed this issue during today's GCST meeting. Let's enable this feature implementation to proceed for GEOS-IT, Merra-2 and GEOS-FP since that information has already been provided. The GCAP implementation could continue to use the older based on U10m and V10m fields with a warning message. |
Name and Institution (Required)
Name: Yuanjian Zhang
Institution: WashU
Describe the update
Use USTAR friction velocity from meteorology instead of calculating from wind at 10m.
Comment out unused codes after this update but keep them following tradition.
Expected changes
Friction velocity is estimated to increase by 3 times. Dust emission is estimated to increase by 30 times, since dust emission is proportional to cube of friction velocity. Therefore, dust scaling factor needs to be recalculated. Following Meng, J.,et al, 2021, global annual dust emission for 2016 is scaled to 2000Tg.
Scaling factor
GEOS-IT(0.5x0.625): 2.6421e-05
MERRA-2(0.5x0.625): 3.8891e-05
GEOS-FP(0.25x0.3125): 3.0655e-05
Related Github Issue
#278