cloud cover change associated with xu_randall #789

RuiyuSun · 2021-12-01T12:49:25Z

Modify the xu/randall cloud cover (progcld6) to have more cloud cover

Fixes #765

Associated PRs:

NOAA-EMC/fv3atm#443
ufs-community/ufs-weather-model#962

climbfuji · 2021-12-02T03:01:43Z

Forgive me my ignorance, but didn't @AnningCheng-NOAA find that the new Thompson cloud cover scheme in PR #781 produces too much cloud cover? Does this mean that once #781 is merged, this PR won't be needed?

As per our offline discussion, my understanding was wrong - this PR can be merged with or after #781.

yangfanglin · 2021-12-02T03:05:51Z

Dom, We need to make a decision in December whether or not to include Thompson MP in the UFS Prototype 8. After discussing with Greg, Ruiyu and a few other developers we came up with two plans. Plan A for the short term: include in P8 all but the ice_RHc updates Ruiyu had made before, and additional tuning of Xu&Randall to improve cloud cover and radiative fluxes. So far the results are encouraging. Plan B: use Greg's branch and tune the parameters to improve cloud cover and radiative fluxes. This will take some time to complete. Once plan B is ready, it will be used to replace plan A in the prototypes. To conclude, this PR is still needed to meet our short-term goals. Thanks Fanglin

…

On Wed, Dec 1, 2021 at 8:42 PM Dom Heinzeller ***@***.***> wrote: Forgive me my ignorance, but didn't @AnningCheng-NOAA <https://github.com/AnningCheng-NOAA> find that the new Thompson cloud cover scheme in PR #781 <#781> produces too much cloud cover? Does this mean that once #781 <#781> is merged, this PR won't be needed? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#789 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKY5N2KT7BSTSCTL56K56DTUO3FGPANCNFSM5JEMP6QA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- *Fanglin Yang, Ph.D.* *Chief, Model Physics Group* *Modeling and Data Assimilation Branch* *NOAA/NWS/NCEP Environmental Modeling Center* *https://www.emc.ncep.noaa.gov/gmb/wx24fy/fyang/ <https://www.emc.ncep.noaa.gov/gmb/wx24fy/fyang/>*

climbfuji · 2021-12-02T03:08:14Z

Dom, We need to make a decision in December whether or not to include Thompson MP in the UFS Prototype 8. After discussing with Greg, Ruiyu and a few other developers we came up with two plans. Plan A for the short term: include in P8 all but the ice_RHc updates Ruiyu had made before, and additional tuning of Xu&Randall to improve cloud cover and radiative fluxes. So far the results are encouraging. Plan B: use Greg's branch and tune the parameters to improve cloud cover and radiative fluxes. This will take some time to complete. Once plan B is ready, it will be used to replace plan A in the prototypes. To conclude, this PR is still needed to meet our short-term goals. Thanks Fanglin
…

Thanks, Fanglin, this sounds like a good plan. Please let us know if there is something we can do to facilitate the process.

grantfirl

Straightforward changes. Approved.

grantfirl · 2021-12-03T20:50:00Z

@climbfuji Are you going to put this and #781 in a wrapper PR?

climbfuji · 2021-12-03T20:55:27Z

@climbfuji Are you going to put this and #781 in a wrapper PR?

Yes - thanks for the reminder. I will then direct the fv3atm that goes with Greg's changes to the combined PR.

climbfuji · 2021-12-06T21:54:58Z

This PR was pulled into #795. It will be merged (and automatically closed) when #795 gets merged.

SamuelTrahanNOAA

Changes look good. One warning though: when you type "0.49," you get 0.49 in default real kind, not kind_phys. You might be doing calculations at a lower precision than you intended. The rest of the file has the same issue, so I see no reason to deviate from that in this PR's changes.

AnningCheng-NOAA · 2021-12-09T15:41:33Z

approved.

climbfuji · 2021-12-10T02:40:53Z

This PR was pulled into #795. It will be merged (and automatically closed) when #795 gets merged.

Update. This PR was removed again from #795, because it led to crashes with GSL physics when Thompson MP was used (in DEBUG and PROD mode, both with GNU and Intel). These issues need to be addressed separately to not hold up the original PR from @gthompsnWRF.

grantfirl · 2021-12-10T17:05:51Z

I reset the code review approvals due to the crashes so that this will be blocked from being merged as-is (in addition to the added label -- does that actually prevent merging?)

climbfuji

This PR needs to be fixed, because the proposed changes do not work with the GSL physics (segmentation faults in the regression tests). The cnvw_in array (which is in phy_f3d) isn't available for Thompon MP. I am not sure how this worked for @RuiyuSun. Also, this array isn't going into the restart files when Thompson MP is used, regardless of the rest of the physics, therefore breaking restart reproducibility.

@RuiyuSun and I will be working on this PR next week.

…sics into HEAD

… crashes

Cloud cover xr dom

climbfuji · 2021-12-16T13:57:35Z

@RuiyuSun please check carefully that I resolved the merge conflicts in physics/radiation_clouds.f correctly: https://github.com/NCAR/ccpp-physics/pull/789/files/ - thanks!

gthompsnWRF

In general, I have no problems with the changes, but do have some pretty simple questions and ideas to mention.

gthompsnWRF · 2021-12-16T14:16:11Z

physics/radiation_clouds.f

-!     real (kind=kind_phys), parameter :: xrc3 = 200.
-      real (kind=kind_phys), parameter :: xrc3 = 100.
+      real (kind=kind_phys), parameter :: xrc3 = 200.
+!     real (kind=kind_phys), parameter :: xrc3 = 100.


Not sure why we need to keep the old value of xrc3 in a comment. From doing git differences of versions, this can be detected pretty easily. Furthermore, this constant is used (I think) in only a single line of code, so is it really needed as a declared variable at all?

gthompsnWRF · 2021-12-16T14:21:15Z

physics/radiation_clouds.f

@@ -3065,6 +3065,7 @@ subroutine progcld6                                               &
        do k = 1, NLAY
          do i = 1, IX
            clwf(i,k) = clw(i,k,ntcw) +  clw(i,k,ntiw) + clw(i,k,ntsw)
+     &      + clw(i,k,ntrw) + cnvw(i,k)


I don't really agree that rain water should be used in a sum of so-called "cloud" species. Rain is not radiatively important to RRTMG anyway. What is being permitted here is the idea that rain contributes to cloud fraction. For what reason? I can describe a scenario with fast moving upper clouds causing snow that falls to a melting layer becoming rain and having zero cloud above it. While you might say this is splitting hairs, as we get to higher and higher resolution that would imply rain with no cloud above still produces a non-zero cloud cover. I am just saying that rain can horizontally separate from the clouds that produced it (in the vertical sense).

gthompsnWRF · 2021-12-16T14:21:46Z

physics/radiation_clouds.f

@@ -3091,8 +3092,9 @@ subroutine progcld6                                               &
            cwp(i,k) = max(0.0, clw(i,k,ntcw) * gfac * delp(i,k))
            cip(i,k) = max(0.0, clw(i,k,ntiw) * gfac * delp(i,k))
            crp(i,k) = max(0.0, clw(i,k,ntrw) * gfac * delp(i,k))
-            csp(i,k) = max(0.0, (clw(i,k,ntsw)+clw(i,k,ntgl)) *         &
-     &                  gfac * delp(i,k))
+!           csp(i,k) = max(0.0, (clw(i,k,ntsw)+clw(i,k,ntgl)) *         &


Graupel was removed, thank you. But I still wonder why we need the old line of code commented out rather than eliminated.

gthompsnWRF · 2021-12-16T14:22:50Z

physics/radiation_clouds.f

@@ -3125,27 +3127,31 @@ subroutine progcld6                                               &
        clwmin = 0.0
        do k = 1, NLAY-1
        do i = 1, IX
-          clwt = 1.0e-6 * (plyr(i,k)*0.001)
+!         clwt = 1.0e-6 * (plyr(i,k)*0.001)


Same comment. Do we now need multiple commented out lines when a single scalar value has been changed? I suggest deleting commented lines.

gthompsnWRF · 2021-12-16T14:29:05Z

physics/radiation_clouds.f

-            if (.not. lmfshal) then
-              tem1  = min(max(sqrt(sqrt(onemrh*qstl(i,k))),0.0001),1.0)
-              tem1  = 2000.0 / tem1
+            if(rhly(i,k) > 1.) then


I will make a strong point of caution here. This line is saying that if Relative Humidity is greater than 100%, then we have 100% cloud coverage. Firstly, in water cloud conditions RH is likely to be machine-precision close to 1.0, but not assured. That might mean RH is numerically 0.99* (many digits of 9). I didn't run through the rest of the lines of code, but if cloud water mixing ratio is something like 1.E-3 (kg/kg; that's pretty big), does that assure a cloud fraction = 100%?

Furthermore, what RH function is being used here below T=0C? Is it the RH with respect to ice or some blend of RH-water and RH-ice? My concern here is that when RH-ice is 101% then result is 100% cloudy, but I can assure you that the real atmosphere is not always cloudy in those conditions. It often takes a lot of ice supersaturation to nucleate the ice crystals. So in the circumstance of RH-ice=105% and the explicit microphysics is making no ice (due to reason above), then do you really want to assume 100% cloudy? If this variable rhly is strictly RH-water, then my point is moot.

gthompsnWRF · 2021-12-16T14:30:58Z

physics/radiation_clouds.f

-              tem1  = min(max(sqrt(sqrt(onemrh*qstl(i,k))),0.0001),1.0)
-              tem1  = 2000.0 / tem1
+            if(rhly(i,k) > 1.) then
+              cldtot(i,k) = 1.
            else


From the prev version of progcld6, you removed the test of whether or not lmfshal is true/false, so there is a different calculation of tem1. I am noting it just to be sure that was intended and not skipped by mistake.

I hope that this was not a result of me resolving the merge conflicts incorrectly. All other merge conflicts I think I got right.

physics/radiation_clouds.f

grantfirl · 2021-12-16T23:19:19Z

@RuiyuSun @yangfanglin Since there is a high priority to get this merged, we have this at the top of the ccpp-physics queue. Do you want to merge this PR as-is or pause to make changes from Greg's comments? A third alternative is to merge as-is and add issues to followup with later, of course.

RuiyuSun · 2021-12-17T01:47:48Z

@grantf ***@***.***> I would like to merge it as it is now and will test the change that Greg suggested.

…

On Thu, Dec 16, 2021 at 6:19 PM Grant Firl ***@***.***> wrote: @RuiyuSun <https://github.com/RuiyuSun> @yangfanglin <https://github.com/yangfanglin> Since there is a high priority to get this merged, we have this at the top of the ccpp-physics queue. Do you want to merge this PR as-is or pause to make changes from Greg's comments? A third alternative is to merge as-is and add issues to followup with later, of course. — Reply to this email directly, view it on GitHub <#789 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKFOCLCO7RZ7SVTGSDKDUX3URJXYHANCNFSM5JEMP6QA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>

-- Ruiyu Sun, PhD *IMSG* at NOAA/NWS/NCEP/EMC 5830 University Research Ct., Rm. 2097 College Park, MD 20740 ***@***.*** ***@***.***> 301-683-3787

yangfanglin · 2021-12-17T05:34:06Z

@RuiyuSun Ruiyu, please remove those commented lines. Some of the scientific questions Greg raised can be addressed later. I think in GFS RH is blended for temperatures between -25 or lower and 0. Would it be possible to make a couple of back to the envelop plots to show the GFS version of Xu&Randall cloud fraction as a function of RH and q (both water and ice), similar to Figs 2 and 3 found in https://armweb0-stg.ornl.gov/publications/proceedings/conf09/extended_abs/lazarus1_sm.pdf ? Xu&Randall scheme was developed with "large-scale" grid averaged water and ice condensates and "large-scale" grid average RH.

RuiyuSun · 2021-12-17T15:58:40Z

@fanglin Yang - NOAA Federal ***@***.***> we can certainly test but I can't open the embedded figure. Comments removed and if block is added back into the routine.

…

On Fri, Dec 17, 2021 at 12:35 AM Fanglin Yang ***@***.***> wrote: @RuiyuSun <https://github.com/RuiyuSun> Ruiyu, please remove those commented lines. Some of the scientific questions Greg raised can be addressed later. I think in GFS RH is blended for temperatures between -25 or lower and 0. Would it be possible to make a couple of back to the envelop plots to show the GFS version of Xu&Randall cloud fraction as a function of RH and q (both water and ice), similar to Figs 2 and 3 found in chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/viewer.html?pdfurl=https%3A%2F% 2Farmweb0-stg.ornl.gov%2Fpublications%2Fproceedings%2Fconf09%2Fextended_abs%2Flazarus1_sm.pdf&clen=365191&chunk=true ? Xu&Randall scheme was developed with "large-scale" grid averaged water and ice condensates and "large-scale" grid average RH. — Reply to this email directly, view it on GitHub <#789 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKFOCLGFIPYZ2D2YIJYDAZTURLDVRANCNFSM5JEMP6QA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>

-- Ruiyu Sun, PhD *IMSG* at NOAA/NWS/NCEP/EMC 5830 University Research Ct., Rm. 2097 College Park, MD 20740 ***@***.*** ***@***.***> 301-683-3787

Update cloud_cover_xr with latest change from NCAR ccpp-physics main

yangfanglin · 2021-12-17T16:04:19Z

@RuiyuSun Try this link https://armweb0-stg.ornl.gov/publications/proceedings/conf09/extended_abs/lazarus1_sm.pdf . I will send you the PDF file by email as well.

cloud cover change associated with xu_randall

473eaf6

RuiyuSun requested review from climbfuji, llpcarson and mkavulich as code owners December 1, 2021 12:49

climbfuji mentioned this pull request Dec 3, 2021

Cloud radiation mp NOAA-EMC/fv3atm#424

Closed

grantfirl approved these changes Dec 3, 2021

View reviewed changes

climbfuji mentioned this pull request Dec 6, 2021

Wrapper PR for #781 (New Thompson cloud fraction) and #800 (Remove Laurie from CODEOWNERS) #795

Merged

climbfuji approved these changes Dec 6, 2021

View reviewed changes

SamuelTrahanNOAA approved these changes Dec 7, 2021

View reviewed changes

climbfuji added bug do not merge labels Dec 10, 2021

grantfirl requested review from grantfirl, climbfuji and SamuelTrahanNOAA December 10, 2021 17:04

climbfuji requested changes Dec 10, 2021

View reviewed changes

climbfuji added 2 commits December 15, 2021 08:35

Merge branch 'cloud_cover_xr' of https://github.com/ruiyusun/ccpp-phy…

cc6b5a6

…sics into HEAD

Bug fixes for getting convectice cloud water into Thompson MP without…

39954f5

… crashes

This was referenced Dec 15, 2021

CCPP cloud cover change for Thompson MP associated with Xu Randall NOAA-EMC/fv3atm#443

Merged

CCPP cloud cover change for Thompson MP associated with Xu Randall, several bug fixes for regression testing scripts ufs-community/ufs-weather-model#962

Merged

Merge pull request #2 from climbfuji/cloud_cover_xr_dom

890df49

Cloud cover xr dom

climbfuji removed the bug label Dec 16, 2021

climbfuji removed the do not merge label Dec 16, 2021

climbfuji requested review from gthompsnWRF and yangfanglin December 16, 2021 14:03

gthompsnWRF requested changes Dec 16, 2021

View reviewed changes

yangfanglin approved these changes Dec 17, 2021

View reviewed changes

climbfuji and others added 2 commits December 17, 2021 07:28

Merge branch 'main' of https://github.com/NCAR/ccpp-physics into HEAD

d1bac25

remove comments and add back the missing if (.not. lmfshal) then block

fe1ba70

Merge pull request #3 from climbfuji/cloud_cover_xr_dom_20211217

5600c09

Update cloud_cover_xr with latest change from NCAR ccpp-physics main

climbfuji approved these changes Dec 20, 2021

View reviewed changes

climbfuji merged commit 84654b7 into NCAR:main Dec 20, 2021

climbfuji mentioned this pull request Dec 29, 2021

cloud cover #788

Closed

RuiyuSun mentioned this pull request Jan 5, 2022

Update Prototype Test suites P7->P8 ufs-community/ufs-weather-model#940

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

cloud cover change associated with xu_randall #789

cloud cover change associated with xu_randall #789

RuiyuSun commented Dec 1, 2021 •

edited by climbfuji

Loading

climbfuji commented Dec 2, 2021

yangfanglin commented Dec 2, 2021 via email

climbfuji commented Dec 2, 2021

grantfirl left a comment

grantfirl commented Dec 3, 2021

climbfuji commented Dec 3, 2021

climbfuji commented Dec 6, 2021

SamuelTrahanNOAA left a comment

AnningCheng-NOAA commented Dec 9, 2021

climbfuji commented Dec 10, 2021

grantfirl commented Dec 10, 2021 •

edited

Loading

climbfuji left a comment

climbfuji commented Dec 16, 2021

gthompsnWRF left a comment

gthompsnWRF Dec 16, 2021

gthompsnWRF Dec 16, 2021

gthompsnWRF Dec 16, 2021

gthompsnWRF Dec 16, 2021

gthompsnWRF Dec 16, 2021

gthompsnWRF Dec 16, 2021

climbfuji Dec 16, 2021

grantfirl commented Dec 16, 2021

RuiyuSun commented Dec 17, 2021 via email

yangfanglin commented Dec 17, 2021 •

edited

Loading

RuiyuSun commented Dec 17, 2021 via email

yangfanglin commented Dec 17, 2021

cloud cover change associated with xu_randall #789

cloud cover change associated with xu_randall #789

Conversation

RuiyuSun commented Dec 1, 2021 • edited by climbfuji Loading

climbfuji commented Dec 2, 2021

yangfanglin commented Dec 2, 2021 via email

climbfuji commented Dec 2, 2021

grantfirl left a comment

Choose a reason for hiding this comment

grantfirl commented Dec 3, 2021

climbfuji commented Dec 3, 2021

climbfuji commented Dec 6, 2021

SamuelTrahanNOAA left a comment

Choose a reason for hiding this comment

AnningCheng-NOAA commented Dec 9, 2021

climbfuji commented Dec 10, 2021

grantfirl commented Dec 10, 2021 • edited Loading

climbfuji left a comment

Choose a reason for hiding this comment

climbfuji commented Dec 16, 2021

gthompsnWRF left a comment

Choose a reason for hiding this comment

gthompsnWRF Dec 16, 2021

Choose a reason for hiding this comment

gthompsnWRF Dec 16, 2021

Choose a reason for hiding this comment

gthompsnWRF Dec 16, 2021

Choose a reason for hiding this comment

gthompsnWRF Dec 16, 2021

Choose a reason for hiding this comment

gthompsnWRF Dec 16, 2021

Choose a reason for hiding this comment

gthompsnWRF Dec 16, 2021

Choose a reason for hiding this comment

climbfuji Dec 16, 2021

Choose a reason for hiding this comment

grantfirl commented Dec 16, 2021

RuiyuSun commented Dec 17, 2021 via email

yangfanglin commented Dec 17, 2021 • edited Loading

RuiyuSun commented Dec 17, 2021 via email

yangfanglin commented Dec 17, 2021

RuiyuSun commented Dec 1, 2021 •

edited by climbfuji

Loading

grantfirl commented Dec 10, 2021 •

edited

Loading

yangfanglin commented Dec 17, 2021 •

edited

Loading