cam6_3_155: Update CLUBB and SILHS externals

[Katetc]

New code with some adjustments for the TAUs tuning options.

Could be a place to address or discuss #953

Possibly another UWM CLUBB branch tag coming in the next week or so, but no changes to CAM code will be required for that, so we should be able to start the code reviews now. This does change answers for all CAM6 and cam_dev compsets. I made a 1 year comparison run against cam6_3_145 to see the impact of the new external and it seems low:
Diagnostics Here

@bstephens82 would be good to keep abreast here.

Closes #956
Closes #971
Closes #953

[bstephens82]

Based on the diagnostics, which are very similar before and after, I would guess the differences are due only to order-of-operations changes. I am checking with @vlarson to make sure that's right. There are physics changes in the taus code with this new external, but those should not appear in the tests if clubb_l_diag_Lscale_from_tau=F, as it is on the trunk.

[vlarson]

Based on the diagnostics, which are very similar before and after, I would guess the differences are due only to order-of-operations changes. I am checking with @vlarson to make sure that's right. There are physics changes in the taus code with this new external, but those should not appear in the tests if clubb_l_diag_Lscale_from_tau=F, as it is on the trunk.

Which two CLUBB tags are showing differences?

[bstephens82]

Based on the diagnostics, which are very similar before and after, I would guess the differences are due only to order-of-operations changes. I am checking with @vlarson to make sure that's right. There are physics changes in the taus code with this new external, but those should not appear in the tests if clubb_l_diag_Lscale_from_tau=F, as it is on the trunk.

Which two CLUBB tags are showing differences?

I believe it's just the previous tag,

clubb_4ncar_20221129_59cb19f_20230330_branchtag

versus the newer tag from 11-15-23:

clubb_4ncar_20231115_5406350

[vlarson]

Based on the diagnostics, which are very similar before and after, I would guess the differences are due only to order-of-operations changes. I am checking with @vlarson to make sure that's right. There are physics changes in the taus code with this new external, but those should not appear in the tests if clubb_l_diag_Lscale_from_tau=F, as it is on the trunk.

Which two CLUBB tags are showing differences?

I believe it's just the previous tag,

clubb_4ncar_20221129_59cb19f_20230330_branchtag

versus the newer tag from 11-15-23:

clubb_4ncar_20231115_5406350

If the comparison is between the newer tag and a tag as old as 2022-11-29, then yes, there are round-off level code changes, such as

larson-group/clubb_release@711714c

It sounds like Kate's comparison runs are demonstrating that there are no bigger changes than round off.

[nusbaume]

Looks good, but I do have some questions and (hopefully minor) change requests.

[cacraigucar]

Mainly documentation clarifications for the namelists. Since these go into the CAM namelist web page, they need to be informative for users.

[adamrher]

Hi Kate. My only comments are on the new CAM namelists. I have a specific question about the clubb_l_diffuse_rtm_and_thlm namelist below. But for the others, I want to confirm whether any of these are turning on new features that weren't on in the old external, or changing the value of parameters w.r.t the old external).

These are new logical namelists:

<clubb_l_calc_thlp2_rad > .true. </clubb_l_calc_thlp2_rad>
<clubb_l_fix_w_chi_eta_correlations > .true. </clubb_l_fix_w_chi_eta_correlations>
<clubb_l_tke_aniso > .true. </clubb_l_tke_aniso>
<clubb_l_upwind_xm_ma > .true. </clubb_l_upwind_xm_ma>
<clubb_l_use_precip_frac > .true. </clubb_l_use_precip_frac>

(These are only the new ones that were added and set to .true.. Ignoring the .false. ones).

Here are the other non-logicals:

<clubb_bv_efold > 5.0 </clubb_bv_efold>
<clubb_wpxp_Ri_exp                                > 0.5     </clubb_wpxp_Ri_exp>
<clubb_z_displace                                 > 25.0    </clubb_z_displace>

[nusbaume]

All of my concerns have been resolved now. Thanks!

[cacraigucar]

Still have my question about threading

[Katetc]

Hi Kate. My only comments are on the new CAM namelists. I have a specific question about the clubb_l_diffuse_rtm_and_thlm namelist below. But for the others, I want to confirm whether any of these are turning on new features that weren't on in the old external, or changing the value of parameters w.r.t the old external).

These are new logical namelists:

<clubb_l_calc_thlp2_rad > .true. </clubb_l_calc_thlp2_rad>
<clubb_l_fix_w_chi_eta_correlations > .true. </clubb_l_fix_w_chi_eta_correlations>
<clubb_l_tke_aniso > .true. </clubb_l_tke_aniso>
<clubb_l_upwind_xm_ma > .true. </clubb_l_upwind_xm_ma>
<clubb_l_use_precip_frac > .true. </clubb_l_use_precip_frac>

(These are only the new ones that were added and set to .true.. Ignoring the .false. ones).

Here are the other non-logicals:

<clubb_bv_efold > 5.0 </clubb_bv_efold>
<clubb_wpxp_Ri_exp                                > 0.5     </clubb_wpxp_Ri_exp>
<clubb_z_displace                                 > 25.0    </clubb_z_displace>

I checked that the logicals were all set to the same values in the previous external. The new non-logicals didn't exist in the previous tag, so there's nothing to compare them too. To see the impacts from this new external, there's comparison diagnostics here: https://webext.cgd.ucar.edu/F2000climo/newCLUBBtesting/larson_tag_20231115.katemerge.011124-1252.F2000dev.f09_f09_mg17_1_2_vs_larson_tag_control.cam6_3_145.011124-1252.F2000dev.f09_f09_mg17_1_2/

[adamrher]

I checked that the logicals were all set to the same values in the previous external. The new non-logicals didn't exist in the previous tag, so there's nothing to compare them too. To see the impacts from this new external, there's comparison diagnostics here

Thanks Kate! PR looks good to me.

[bstephens82]

Based on the diagnostics, which are very similar before and after, I would guess the differences are due only to order-of-operations changes. I am checking with @vlarson to make sure that's right. There are physics changes in the taus code with this new external, but those should not appear in the tests if clubb_l_diag_Lscale_from_tau=F, as it is on the trunk.

Resolved.

All comments made by @bstephens82 have been resolved