Helium in TIE-GCM and TIME-GCM
7 messages
SUTTON, ERIC K DR-02 USAF AFMC AFRL/RVBXI <eric.sutton.5@us.af.mil>	Wed, Nov 5, 2014 at 11:30 AM
To: "Stan Solomon (stans@ucar.edu)" <stans@ucar.edu>, "foster@ucar.edu" <foster@ucar.edu>, "wbwang >> Wen-Bin Wang (wbwang@ucar.edu)" <wbwang@ucar.edu>
Cc: "Jeffrey P Thayer (Jeffrey.Thayer@colorado.edu)" <Jeffrey.Thayer@colorado.edu>
Hi Stan, Ben and Wenbin,

Jeff mentioned to me just the other day that you have been working to get
helium running in TIE-GCM and TIME-GCM trunk versions. After looking through
the change-logs last night, I see a lot has been accomplished since the last
commit to the tiegcm_he branch. Since I only received emails for the tiegcm_he
branch commits, I missed quite a bit and had assumed the helium work was in a
holding pattern. I'm really glad to see so much progress has been made, and
especially that the code has made it into both TIE and TIME models. If I can
help in any way please let me know (e.g. expanding alfa from 2x2 to 3x3 in
minor.F).

I have a couple of edits to make in laplacian.F and to the spectral
coefficients to fix an error I originally made. Among other things, this will
fix the large magnitudes in exospheric flux (wave-1 in longitude) right around
the poles that I had originally brushed off as being caused by some other
numerical issue. This change may require further adjustment to the degree of
truncation in the spherical harmonic expansion in laplacian.F to make sure the
model is stable. I had intended to apply these updates to the tiegcm_he
branch, but it's probably better to apply them directly to the TIEGCM and
TIMEGCM trunks, in which case I can hand off the code/coefficients to Ben once
I've run some initial tests. Ben, if that's ok, I'll checkout the latest
TIE-GCM trunk and apply the modifications to that version and email you the
changes.

Also, I sent Jeff the paper draft for the helium modeling work, and once I
incorporate any suggestions he has, I'll send it your way as well.

Thanks and best regards,
-Eric

---------------------------
Eric K. Sutton, PhD
Research Physicist
AFRL/RVBXI
(505) 846-7846


-----Original Message-----
From: foster@ucar.edu [mailto:foster@ucar.edu]
Sent: Monday, February 03, 2014 11:02 AM
To: esutton@ucar.edu
Subject: SVN: tgcm [1012] tiegcm/branches/tiegcm_he:

Revision: 1012
Author:   foster
Date:     2014-02-03 11:01:36 -0700 (Mon, 03 Feb 2014)
Log Message:
-----------
tiegcm/branches/tiegcm_he:
 - Update this branch from the trunk, including all changes to the trunk
   since this branch was started (r964). This was done w/ the following
   merge command, executed from an up-to-date working copy of the branch:
   svn merge -r964:HEAD $SVN/tiegcm/trunk .
   There were no conflicts. This is in preparation for merging the helium
   branch to the trunk.

Modified Paths:
--------------
    tiegcm/branches/tiegcm_he/scripts/Make.intel_ys
    tiegcm/branches/tiegcm_he/src/advance.F
    tiegcm/branches/tiegcm_he/src/cism_coupling.F
    tiegcm/branches/tiegcm_he/src/cism_intercomm.F
    tiegcm/branches/tiegcm_he/src/filter.F
    tiegcm/branches/tiegcm_he/src/gpi.F
    tiegcm/branches/tiegcm_he/src/gswm.F
    tiegcm/branches/tiegcm_he/src/input.F
    tiegcm/branches/tiegcm_he/src/minor.F
    tiegcm/branches/tiegcm_he/src/newton.F
    tiegcm/branches/tiegcm_he/src/oplus.F
    tiegcm/branches/tiegcm_he/src/output.F
    tiegcm/branches/tiegcm_he/src/qjoule.F
    tiegcm/branches/tiegcm_he/src/tgcm.F
    tiegcm/branches/tiegcm_he/src/wei05sc.F

Property Changed:
----------------
    tiegcm/branches/tiegcm_he/


		smime.p7s
6K
Ben Foster <foster@ucar.edu>	Wed, Nov 5, 2014 at 12:33 PM
To: "SUTTON, ERIC K DR-02 USAF AFMC AFRL/RVBXI" <eric.sutton.5@us.af.mil>
Cc: "Stan Solomon (stans@ucar.edu)" <stans@ucar.edu>, "wbwang >> Wen-Bin Wang (wbwang@ucar.edu)" <wbwang@ucar.edu>, "Jeffrey P Thayer (Jeffrey.Thayer@colorado.edu)" <Jeffrey.Thayer@colorado.edu>
Eric,

Sorry you were not receiving notifications of commits to the trunk versions of tiegcm and timegcm. Let me know if you would like to receive email of commits to the trunks, and I will have CISL add you to the lists.

Yes, please send me your modifications to Laplacian (same for both models?), and I will test and commit to the repository.  BTW, I'm not sure what your attachment smime.p7s is - looks like a binary file, my linux desktop cannot open it.

Thanks a lot for working on this, and again, I'm sorry you were not notified about the commits.

--Ben
[Quoted text hidden]
-- 
Ben Foster
National Center for Atmospheric Research (NCAR)
High Altitude Observatory (HAO)
303-497-1595
SUTTON, ERIC K DR-02 USAF AFMC AFRL/RVBXI <eric.sutton.5@us.af.mil>	Wed, Nov 5, 2014 at 1:33 PM
To: Ben Foster <foster@ucar.edu>
Cc: "wbwang >> Wen-Bin Wang (wbwang@ucar.edu)" <wbwang@ucar.edu>, "Jeffrey P Thayer (Jeffrey.Thayer@colorado.edu)" <Jeffrey.Thayer@colorado.edu>, "Stan Solomon (stans@ucar.edu)" <stans@ucar.edu>
Hi Ben,

Not at all, I was very pleasantly surprised when I saw all the updates.

It would be great to be on future trunk email traffic though, if it's not much
trouble; thanks!

I think the mods should be very similar for both models. But if nothing else,
I'm betting the truncation degree might need to be adjusted up or down for
each. I'll look into it and let you know what I find out.

I think that attachment is my digital signature or something similar, so feel
free to disregard.

Thanks again,
-Eric
[Quoted text hidden]
        (505) 846-7846 <tel:%28505%29%20846-7846>
[Quoted text hidden]

		smime.p7s
6K
Ben Foster <foster@ucar.edu>	Wed, Nov 5, 2014 at 4:30 PM
To: "cislhelp@mail.ucar.edu" <cislhelp@mail.ucar.edu>

Attention: Gary Studwell, re tgcm repository.

Gary,

Please add Eric Sutton (eric.sutton.5@us.af.mil) to email commit notices for the tiegcm and timegcm trunk codes (I believe he is currently being notified only for commits to tiegcm/branches/tiegcm_he).
Thanks,

--Ben
[Quoted text hidden]

		smime.p7s
6K
SUTTON, ERIC K DR-02 USAF AFMC AFRL/RVBXI <eric.sutton.5@us.af.mil>	Mon, Nov 10, 2014 at 10:00 AM
To: Ben Foster <foster@ucar.edu>
Hi Ben,

I've made the changes to the exospheric flux code, but ran into a snag trying
to test it out. I only have PGI fortran installed in my office, and it looks
like this may not be supported after pdynamo revisions due to ESMF libraries.
I'm working toward ordering Intel fortran and setting up ESMF libraries, but I
don't want to hold up the process. Over the weekend, I logged into an HAO
workstation (titanium) and made a few quick 'sanity check' runs with the
modifications, and everything seems to be working fine. I've attached the
changes and here's a quick run-down of the modifications:

- Deleted the data array lpmn which contained (m^2/sin(theta)-n*(n+1))*pmn.
The m^2/sin(theta) term is not needed. Therefore, I replaced lpmn with pmn
(the normalized associated legendre polynomials) and added the
latitude-independent term -n*(n+1) to the synthesis loop within laplacian.F.
- Added compiler IF statement and parameter 'truncdeg' to define the degree of
truncation for the spectral approximation, which depends on grid resolution
and upper boundary pressure level (i.e. z=+7 for TIEGCM vs. z=+5 for TIMEGCM).
Currently set truncdeg=4 for default resolution, and truncdeg=8 for double
res.
- Minor updates to comments in laplacian.F, comp.F, he_coefs_sres.F, and
he_coef0_dres.F.
- Uncommented 2 addiag calls for 'FLX_HE' and 'FLX_ARG' in comp.F to
facilitate testing the stability of tiegcm + helium/exospheric model.
- The 'Fit' and 'Synthesis' sections of laplacian.F each now contain commented
loops (just after the main loops) that are helpful when debugging various
levels of truncation and testing for stability.

I checked out r1103 of TIEGCM and made the above changes to that revision. I
also checked out r1103 of TIMEGCM and as far as I can tell, the modifications
should be exactly the same for TIMEGCM as for TIEGCM. I've only had a chance
to test TIEGCM sres, and only for one short run near solar max. When tuning
'truncdeg', I look in the variable 'FLX_HE' for waves that amplify rather than
diminish over time. If these waves are effective in coupling into 'HE', a
feedback occurs and the solution degrades quickly. I chose solar max because
tiegcm + helium exospheric model is least stable under these conditions. I'm
confident that 'truncdeg=4' is more than adequate for all resolutions, but I
expect that higher degrees will work fine (i.e. 'truncdeg=8') for TIEGCM dres,
yielding more accurate results for the exospheric code. I also expect that the
same will go for TIMEGCM, but I haven't yet tested this.

Thanks for your help,
-Eric

---------------------------
Eric K. Sutton, PhD
Research Physicist
AFRL/RVBXI
(505) 846-7846

-----Original Message-----
From: SUTTON, ERIC K DR-02 USAF AFMC AFRL/RVBXI
Sent: Wednesday, November 05, 2014 1:33 PM
To: 'Ben Foster'
Cc: wbwang >> Wen-Bin Wang (wbwang@ucar.edu); Jeffrey P Thayer
(Jeffrey.Thayer@colorado.edu); Stan Solomon (stans@ucar.edu)
Subject: RE: Helium in TIE-GCM and TIME-GCM

[Quoted text hidden]
[Quoted text hidden]
        (505) 846-7846 <tel:%28505%29%20846-7846>
[Quoted text hidden]

7 attachments
		comp.F
44K
		diff_comp.txt
1K
		diff_laplacian.txt
4K
		he_coef0_dres.F
4479K
		he_coefs_sres.F
1147K
		laplacian.F
7K
		smime.p7s
6K 
