How do I add aux3d?

[ericaligo-NOAA]

This document here(https://ccpp-techdoc.readthedocs.io/en/v5.0.0/ParamSpecificOutput.html#output-of-auxiliary-arrays-from-ccpp) talks about adding diagnostics. I've added a 3D diagnostic to a microphysics scheme called vtcsk(k). Where do I add aux3d(:,:1)? Within a do loop or outside of a do loop? Would it be something like:

 aux3d(:,:,1) = vtcsk(k)
 aux3d(:,:,2) = vtcrk(k)

or within the k loop?
do k = kts, kte
qv(i,k,j) = qv1d(k)
aux3d(:,:,1) = vtcsk(k)
aux3d(:,:,2) = vtcrk(k)
Eric

[mzhangw]

Is the 3D diagnostic vtcsk(i,k) or vtcrk(i,k) by any chance? In this case, you can use:
aux3d(:,:,1)=vtcsk(:,:)
aux3d(:,:,2)=vtcrk(:,:)

[ericaligo-NOAA]

Man, vtcsk is just a function of k in mp_thompson subroutine as are the other fields, but outside of that routine that array is passed to a 3D array, like qv as shown below, that is a function of i,j,k.  My output looks strange, so I suspect what I have below is not right?       j_loop:  do j = j_start, j_end       i_loop:  do i = i_start, i_end          RAINNCV(i,j) = 0. !> - Call mp_thompson()          call mp_thompson(qv1d, qc1d, qi1d, qr1d, qs1d, qg1d, ni1d,     &                       nr1d, nc1d, nwfa1d, nifa1d, t1d, p1d, w1d, dz1d,  &                       pptrain, pptsnow, pptgraul, pptice, &                       rand1, rand2, rand3, &                       kts, kte, dt, i, j,*vtcsk*)          do k = kts, kte             qv(i,k,j) = qv1d(k) *aux3d(i,k,1) = vtcsk(k)*          enddo

…

On 5/6/2021 11:29 PM, mzhangw wrote: Is the 3D diagnostic vtcsk(i,k) or vtcrk(i,k) by any chance? In this case, you can use: aux3d(:,:,1)=vtcsk(:,:) aux3d(:,:,2)=vtcrk(:,:) — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75MICBQDB3HSOOPGS5R3TMNNBHANCNFSM44F5CS4A>.

[mzhangw]

aux2d and aux3d are defined in a column profile with the dimension (1:ncol,#) and (1:ncol,1:nlev,#). I would suggest to pass the vtcsk to an intermediate variable with dimension (i,k,j) in module_mp_thompson.F90, and define aux3d(1:ncol,1:,nlev,1) in thompson driver wrapper mp_thompson.F90.

[ericaligo-NOAA]

Thanks.  aux3d has three places, though, right? aux3d(:,:,1). In your example below, is 1:nlev the vertical levels or the number representing the diagnostic array where nlev  is the number of diagnostics?

…

On 5/7/2021 3:59 PM, mzhangw wrote: axu3d and aux2d is defined in a column profile with the dimension (1:ncol: 1,nlev). I would suggest to pass the vtcsk to something immediate variable with dimension (i,k,j) in module_mp_thompson.F90, and define aux3d(1:ncol,1:,nlev) in thompson driver wrapper mp_thompson.F90. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#650 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75MINTJWMMZIZFVIAA5LTMRBCPANCNFSM44F5CS4A>.

[mzhangw]

Yes, you are correct. # is the number of diagnostics. I updated my previous comments.

[ericaligo-NOAA]

Thanks.  I'll make some changes and see what happens.

…

On 5/7/2021 4:21 PM, mzhangw wrote: Yes, you are correct. # is the number of diagnostics. I updated my previous comments. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#650 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75MLCM4QSZMNDTCEMEKDTMRDT3ANCNFSM44F5CS4A>.

[ligiabernardet]

@ericaligo-NOAA Let us know how it goes and if you are making progress in outputting the Thompson mp diagnostics.

[ericaligo-NOAA]

Thanks.  I would like someone to see what I did to make sure this is correct and the most efficient way to use the aux files.

…

On 5/12/2021 4:46 PM, ligiabernardet wrote: @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know how it goes and if you are making progress in outputting the Thompson mp diagnostics. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#650 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>.

[climbfuji]

Thanks, Eric. You can always put your code changes in a branch so that one can clone it to view the diffs offline, or use github for a visualization.

…

On May 12, 2021, at 2:58 PM, ericaligo-NOAA ***@***.***> wrote: Thanks. I would like someone to see what I did to make sure this is correct and the most efficient way to use the aux files. On 5/12/2021 4:46 PM, ligiabernardet wrote: > > @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know how it > goes and if you are making progress in outputting the Thompson mp > diagnostics. > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#650 (reply in thread)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>. > — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RKD2UTONT6DP7WQUV3TNLTYHANCNFSM44F5CS4A>.

[ericaligo-NOAA]

Could I simply point someone to my path?

…

On 5/12/2021 5:00 PM, Dom Heinzeller wrote: Thanks, Eric. You can always put your code changes in a branch so that one can clone it to view the diffs offline, or use github for a visualization. > On May 12, 2021, at 2:58 PM, ericaligo-NOAA ***@***.***> wrote: > > > Thanks. I would like someone to see what I did to make sure this is > correct and the most efficient way to use the aux files. > > On 5/12/2021 4:46 PM, ligiabernardet wrote: > > > > @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know how it > > goes and if you are making progress in outputting the Thompson mp > > diagnostics. > > > > — > > You are receiving this because you were mentioned. > > Reply to this email directly, view it on GitHub > > <#650 (reply in thread)>, > > or unsubscribe > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>. > > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RKD2UTONT6DP7WQUV3TNLTYHANCNFSM44F5CS4A>. > — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75ML7LD66YRCD52FVWLLTNLT6BANCNFSM44F5CS4A>.

[climbfuji]

Branches are definitely preferred over a directory on machine xyz. No need to come up with a sophisticated commit history for code that won't get merged into the authoritative repository.

…

On May 12, 2021, at 3:07 PM, ericaligo-NOAA ***@***.***> wrote: Could I simply point someone to my path? On 5/12/2021 5:00 PM, Dom Heinzeller wrote: > Thanks, Eric. You can always put your code changes in a branch so that > one can clone it to view the diffs offline, or use github for a > visualization. > > > On May 12, 2021, at 2:58 PM, ericaligo-NOAA ***@***.***> wrote: > > > > > > Thanks. I would like someone to see what I did to make sure this is > > correct and the most efficient way to use the aux files. > > > > On 5/12/2021 4:46 PM, ligiabernardet wrote: > > > > > > @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know > how it > > > goes and if you are making progress in outputting the Thompson mp > > > diagnostics. > > > > > > — > > > You are receiving this because you were mentioned. > > > Reply to this email directly, view it on GitHub > > > > <#650 (reply in thread)>, > > > > or unsubscribe > > > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>. > > > > > — > > You are receiving this because you are subscribed to this thread. > > Reply to this email directly, view it on GitHub > <#650 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AB5C2RKD2UTONT6DP7WQUV3TNLTYHANCNFSM44F5CS4A>. > > > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#650 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ALQ75ML7LD66YRCD52FVWLLTNLT6BANCNFSM44F5CS4A>. > — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RLGVWMJMOLUJDVBLLDTNLUZLANCNFSM44F5CS4A>.

[ericaligo-NOAA]

 Here's a question.  aux3d(:,:1) is an instantaneous field in red.  If i want to average over time, besides specifying that in the namelist, do I need to do something like what I have in bold blue where condensation is instantaneous? *aux3d(:,:,1) = vts1* *aux3d(:,:2) = aux3d(:,:2) + condensation* Eric

…

On 5/12/2021 5:12 PM, Dom Heinzeller wrote: Branches are definitely preferred over a directory on machine xyz. No need to come up with a sophisticated commit history for code that won't get merged into the authoritative repository. > On May 12, 2021, at 3:07 PM, ericaligo-NOAA ***@***.***> wrote: > > > Could I simply point someone to my path? > > On 5/12/2021 5:00 PM, Dom Heinzeller wrote: > > Thanks, Eric. You can always put your code changes in a branch so that > > one can clone it to view the diffs offline, or use github for a > > visualization. > > > > > On May 12, 2021, at 2:58 PM, ericaligo-NOAA ***@***.***> wrote: > > > > > > > > > Thanks. I would like someone to see what I did to make sure this is > > > correct and the most efficient way to use the aux files. > > > > > > On 5/12/2021 4:46 PM, ligiabernardet wrote: > > > > > > > > @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know > > how it > > > > goes and if you are making progress in outputting the Thompson mp > > > > diagnostics. > > > > > > > > — > > > > You are receiving this because you were mentioned. > > > > Reply to this email directly, view it on GitHub > > > > > > <#650 (reply in thread)>, > > > > > > or unsubscribe > > > > > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>. > > > > > > > — > > > You are receiving this because you are subscribed to this thread. > > > Reply to this email directly, view it on GitHub > > <#650 (comment)>, > > or unsubscribe > > <https://github.com/notifications/unsubscribe-auth/AB5C2RKD2UTONT6DP7WQUV3TNLTYHANCNFSM44F5CS4A>. > > > > > > > — > > You are receiving this because you were mentioned. > > Reply to this email directly, view it on GitHub > > <#650 (comment)>, > > or unsubscribe > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML7LD66YRCD52FVWLLTNLT6BANCNFSM44F5CS4A>. > > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RLGVWMJMOLUJDVBLLDTNLUZLANCNFSM44F5CS4A>. > — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75MJ4N5RUADGMO2RCU3DTNLVLRANCNFSM44F5CS4A>.

[climbfuji]

Averaging can only work if you accumulate. The accumulation needs to happen in the code, the averaging is done in the I/O routine. You can of course handle the averaging yourself and not use the namelist. Note also that the averaging in the I/O routine divides by the time interval, so for this to make sense you will want to make sure that the quantity you are accumulating takes that into account, for example: accumulated_rain = accumulated_rain + instantaneous_rain_now * physics_timestep

…

On May 12, 2021, at 3:50 PM, ericaligo-NOAA ***@***.***> wrote: Here's a question. aux3d(:,:1) is an instantaneous field in red. If i want to average over time, besides specifying that in the namelist, do I need to do something like what I have in bold blue where condensation is instantaneous? *aux3d(:,:,1) = vts1* *aux3d(:,:2) = aux3d(:,:2) + condensation* Eric On 5/12/2021 5:12 PM, Dom Heinzeller wrote: > Branches are definitely preferred over a directory on machine xyz. No > need to come up with a sophisticated commit history for code that > won't get merged into the authoritative repository. > > > On May 12, 2021, at 3:07 PM, ericaligo-NOAA ***@***.***> wrote: > > > > > > Could I simply point someone to my path? > > > > On 5/12/2021 5:00 PM, Dom Heinzeller wrote: > > > Thanks, Eric. You can always put your code changes in a branch so > that > > > one can clone it to view the diffs offline, or use github for a > > > visualization. > > > > > > > On May 12, 2021, at 2:58 PM, ericaligo-NOAA ***@***.***> wrote: > > > > > > > > > > > > Thanks. I would like someone to see what I did to make sure this is > > > > correct and the most efficient way to use the aux files. > > > > > > > > On 5/12/2021 4:46 PM, ligiabernardet wrote: > > > > > > > > > > @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know > > > how it > > > > > goes and if you are making progress in outputting the Thompson mp > > > > > diagnostics. > > > > > > > > > > — > > > > > You are receiving this because you were mentioned. > > > > > Reply to this email directly, view it on GitHub > > > > > > > > > <#650 (reply in thread)>, > > > > > > > > > or unsubscribe > > > > > > > > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>. > > > > > > > > > — > > > > You are receiving this because you are subscribed to this thread. > > > > Reply to this email directly, view it on GitHub > > > > <#650 (comment)>, > > > > or unsubscribe > > > > <https://github.com/notifications/unsubscribe-auth/AB5C2RKD2UTONT6DP7WQUV3TNLTYHANCNFSM44F5CS4A>. > > > > > > > > > > — > > > You are receiving this because you were mentioned. > > > Reply to this email directly, view it on GitHub > > > > <#650 (comment)>, > > > > or unsubscribe > > > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML7LD66YRCD52FVWLLTNLT6BANCNFSM44F5CS4A>. > > > > > — > > You are receiving this because you commented. > > Reply to this email directly, view it on GitHub > <#650 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AB5C2RLGVWMJMOLUJDVBLLDTNLUZLANCNFSM44F5CS4A>. > > > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#650 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ALQ75MJ4N5RUADGMO2RCU3DTNLVLRANCNFSM44F5CS4A>. > — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RKLWDBD7G3OKQDBSS3TNLZZXANCNFSM44F5CS4A>.

[ericaligo-NOAA]

The way I used to write out the mp diagnostics was to write out the delta in the process not the rate (condensation amount/dt), so I didn't multiply by physics_timestep, but I think I understand how this works now.  In fact, I wrote out the accumulated process over the entire 1 hour, which I believe, correct me if I'm wrong, would be using the lines in blue while specifying false in the namelist for averaging?  Averaging is fine too, but it's good to know I can simply write out the accumulation.

…

On 5/12/2021 5:56 PM, Dom Heinzeller wrote: Averaging can only work if you accumulate. The accumulation needs to happen in the code, the averaging is done in the I/O routine. You can of course handle the averaging yourself and not use the namelist. Note also that the averaging in the I/O routine divides by the time interval, so for this to make sense you will want to make sure that the quantity you are accumulating takes that into account, for example: *accumulated_rain = accumulated_rain + instantaneous_rain_now* * physics_timestep > On May 12, 2021, at 3:50 PM, ericaligo-NOAA ***@***.***> wrote: > > > Here's a question. aux3d(:,:1) is an instantaneous field in red. If > i want to average over time, besides specifying that in the namelist, do > I need to do something like what I have in bold blue where condensation > is instantaneous? > > *aux3d(:,:,1) = vts1* > > *aux3d(:,:2) = aux3d(:,:2) + condensation* > > Eric > On 5/12/2021 5:12 PM, Dom Heinzeller wrote: > > Branches are definitely preferred over a directory on machine xyz. No > > need to come up with a sophisticated commit history for code that > > won't get merged into the authoritative repository. > > > > > On May 12, 2021, at 3:07 PM, ericaligo-NOAA ***@***.***> wrote: > > > > > > > > > Could I simply point someone to my path? > > > > > > On 5/12/2021 5:00 PM, Dom Heinzeller wrote: > > > > Thanks, Eric. You can always put your code changes in a branch so > > that > > > > one can clone it to view the diffs offline, or use github for a > > > > visualization. > > > > > > > > > On May 12, 2021, at 2:58 PM, ericaligo-NOAA ***@***.***> wrote: > > > > > > > > > > > > > > > Thanks. I would like someone to see what I did to make sure this is > > > > > correct and the most efficient way to use the aux files. > > > > > > > > > > On 5/12/2021 4:46 PM, ligiabernardet wrote: > > > > > > > > > > > > @ericaligo-NOAA <https://github.com/ericaligo-NOAA> Let us know > > > > how it > > > > > > goes and if you are making progress in outputting the Thompson mp > > > > > > diagnostics. > > > > > > > > > > > > — > > > > > > You are receiving this because you were mentioned. > > > > > > Reply to this email directly, view it on GitHub > > > > > > > > > > > > <#650 (reply in thread)>, > > > > > > > > > > > > or unsubscribe > > > > > > > > > > > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML4BKCMZECUNZMEQVTTNLSKDANCNFSM44F5CS4A>. > > > > > > > > > > > — > > > > > You are receiving this because you are subscribed to this thread. > > > > > Reply to this email directly, view it on GitHub > > > > > > <#650 (comment)>, > > > > > > or unsubscribe > > > > > > <https://github.com/notifications/unsubscribe-auth/AB5C2RKD2UTONT6DP7WQUV3TNLTYHANCNFSM44F5CS4A>. > > > > > > > > > > > > > — > > > > You are receiving this because you were mentioned. > > > > Reply to this email directly, view it on GitHub > > > > > > <#650 (comment)>, > > > > > > or unsubscribe > > > > > > <https://github.com/notifications/unsubscribe-auth/ALQ75ML7LD66YRCD52FVWLLTNLT6BANCNFSM44F5CS4A>. > > > > > > > — > > > You are receiving this because you commented. > > > Reply to this email directly, view it on GitHub > > <#650 (comment)>, > > or unsubscribe > > <https://github.com/notifications/unsubscribe-auth/AB5C2RLGVWMJMOLUJDVBLLDTNLUZLANCNFSM44F5CS4A>. > > > > > > > — > > You are receiving this because you were mentioned. > > Reply to this email directly, view it on GitHub > > <#650 (comment)>, > > or unsubscribe > > <https://github.com/notifications/unsubscribe-auth/ALQ75MJ4N5RUADGMO2RCU3DTNLVLRANCNFSM44F5CS4A>. > > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RKLWDBD7G3OKQDBSS3TNLZZXANCNFSM44F5CS4A>. > — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75MKIGRQ7S4M5BAJG2CLTNL2RVANCNFSM44F5CS4A>.

[ericaligo-NOAA]

Would that involve creating a fork and then pushing, and all of that wonderful stuff that I do about once a year? :)

…

On 5/12/2021 5:12 PM, Dom Heinzeller wrote: Branches are definitely preferred over a directory on machine xyz

[climbfuji]

Yes, and the best thing about it is that you will do it twice this year :-)

…

On May 12, 2021, at 4:18 PM, ericaligo-NOAA ***@***.***> wrote: Would that involve creating a fork and then pushing, and all of that wonderful stuff that I do about once a year? :) On 5/12/2021 5:12 PM, Dom Heinzeller wrote: > Branches are definitely preferred over a directory on machine xyz — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RLWNQTU4GSZNIIIUUDTNL5C5ANCNFSM44F5CS4A>.

[ericaligo-NOAA]

Don't know what went wrong. Hera:/scratch2/NCEPDEV/fv3-cam/save/Eric.Aligo/ufs-weather-model.git_tho_diagnostics/FV3>git branch * (HEAD detached at 345915a)   develop Hera:/scratch2/NCEPDEV/fv3-cam/save/Eric.Aligo/ufs-weather-model.git_tho_diagnostics/FV3>git checkout tho_diagnostics error: pathspec 'tho_diagnostics' did not match any file(s) known to git.

…

On 5/12/2021 6:27 PM, Dom Heinzeller wrote: Yes, and the best thing about it is that you will do it twice this year :-) > On May 12, 2021, at 4:18 PM, ericaligo-NOAA ***@***.***> wrote: > > > Would that involve creating a fork and then pushing, and all of that > wonderful stuff that I do about once a year? :) > > On 5/12/2021 5:12 PM, Dom Heinzeller wrote: > > Branches are definitely preferred over a directory on machine xyz > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB5C2RLWNQTU4GSZNIIIUUDTNL5C5ANCNFSM44F5CS4A>. > — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#650 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALQ75MIMUISPMKX6HXRXMHLTNL6GBANCNFSM44F5CS4A>.