Error message when outputting available tendencies #710
Replies: 8 comments 5 replies
-
|
@SamuelTrahanNOAA do you have time to look into this? @ericaligo-NOAA can you make sure that the error is related to the 3d diagnostic tendency namelist options, and not something else? |
Beta Was this translation helpful? Give feedback.
-
|
The You should see the string "dtend_select" several places in FV3/ccpp/data/GFS_typedefs.F90 In general, if you update your FV3 or CCPP code, you need to generate a new test case using the initialization system. Other updates (unrelated to this) over the past few months have added and removed namelist settings and input files. An easy way to get a new case, for immediate debugging purposes, is to grab the working directory of one of the regression tests, and then tune it to your needs. Then compare it to the original to see what is different. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for that info. I do not see dtend_select in GFS_typedefs.F90. A
git log in /ufs-weather-model shows july 16 as the last entry.
Eric
…On 7/27/2021 10:41 AM, Samuel Trahan (NOAA contractor) wrote:
The |dtend_select| is a new namelist option added to main and develop
recently. Does it exist in your version?
You should see the string "dtend_select" several places in
FV3/ccpp/data/GFS_typedefs.F90
In general, if you update your FV3 or CCPP code, you need to generate
a new test case using the initialization system. Other updates
(unrelated to this) over the past few months have added and removed
namelist settings and input files. An easy way to get a new case, for
immediate debugging purposes, is to grab the working directory of one
of the regression tests, and then tune it to your needs. Then compare
it to the original to see what is different.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#710 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MJDSFIRWXVDI2RDTMLTZ3ARRANCNFSM5BBCGXLA>.
|
Beta Was this translation helpful? Give feedback.
-
|
The feature was merged on the 21st. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks!
…On 7/27/2021 10:46 AM, Samuel Trahan (NOAA contractor) wrote:
The feature was merged on the 21st.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#710 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MPOLSRAHFUYXMKH67DTZ3BEFANCNFSM5BBCGXLA>.
|
Beta Was this translation helpful? Give feedback.
-
|
Does one know where to find the old instructions on adding the available
diagnostics (non physics, temperature diagnostics)? I can't recall what
I need to add to the diag_table, if anything, and what changes are all
involved in input.nml. Online, it looks like the new instructions for
the latest commit are there, but since I don't plan on updating my codes
anytime soon, it would be great if the previous method for outputting
diagnostics was online as well.
Eric
…On 7/27/2021 10:46 AM, Samuel Trahan (NOAA contractor) wrote:
The feature was merged on the 21st.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#710 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MPOLSRAHFUYXMKH67DTZ3BEFANCNFSM5BBCGXLA>.
|
Beta Was this translation helpful? Give feedback.
-
|
I'm unaware of any such documentation. I figured it out by reading through the source code. Generally:
Things get more complicated if you want it to be part of coupling, restarting, or inline posting. |
Beta Was this translation helpful? Give feedback.
-
|
A couple of months ago, I followed instructions that were posted under
section 10 here.
<https://ccpp-techdoc.readthedocs.io/en/latest/ParamSpecificOutput.html>
They described a way for adding the physics+non physics related
tendencies to the history files. Those instructions have been updated
and did not require the use of dtend*. Perhaps @DomHeinzeller or @grantfirl might recall
those instructions.
Eric
…On 7/27/2021 12:38 PM, Samuel Trahan (NOAA contractor) wrote:
I'm unaware of any such documentation. I figured it out by reading
through the source code. Generally:
1. FV3/ccpp/data/GFS_typedefs.F90 = declare, allocate, and initialize
fields and namelist variables
2. FV3/ccpp/driver/GFS_diagnostics.F90 = connection to diag_table
3. FV3/ccpp/suites/*.xml = if you need to add schemes to suites for
your diagnostics (like the new phys_tend)
4. FV3/ccpp/physics/physics/ = put the diagnostic code here. The
interstitials are generics are good places if you want it to be
suite-independent.
Things get more complicated if you want it to be part of coupling,
restarting, or inline posting.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#710 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MO344KZK6GZ5BEBSQTTZ3OHRANCNFSM5BBCGXLA>.
|
Beta Was this translation helpful? Give feedback.
-
|
The |
Beta Was this translation helpful? Give feedback.
-
|
Yes, I was using the d*3dt arrays. I cloned the latest code less than 2
weeks ago and don't plan on updating for a while. What needs to be
done is for the original instructions for the d*3dt arrays to be added
back online with a note that those instructions pertain to code prior to
a certain date. After a year you can remove those instructions.
…On 7/27/2021 12:53 PM, Samuel Trahan (NOAA contractor) wrote:
The |dtend| and |dtidx| arrays replace the various |d*3dt| arrays. The
aux arrays haven't changed (section 10.3). If you were using |d*3dt|
then you will have to switch to |dtend|.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#710 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MNJSZCFO5S4ZIFE3FDTZ3P7XANCNFSM5BBCGXLA>.
|
Beta Was this translation helpful? Give feedback.
-
|
Please file a ticket in the ccpp-doc issue page. The documentation you want still exists in an earlier version of the documentation repository (hash 3e9224c, I think) so the website would have to be updated... somehow... to display older versions. |
Beta Was this translation helpful? Give feedback.
-
|
Great. Thanks!
…On 7/27/2021 1:03 PM, Samuel Trahan (NOAA contractor) wrote:
Please file a ticket in the ccpp-doc issue page. The documentation you
want still exists in an earlier version of the documentation
repository (hash 3e9224c, I think) so the website would have to be
updated... somehow... to display older versions.
https://github.com/NCAR/ccpp-doc <https://github.com/NCAR/ccpp-doc>
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#710 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MITLYZJQFP6DLJVX7DTZ3RE5ANCNFSM5BBCGXLA>.
|
Beta Was this translation helpful? Give feedback.
-
|
@ericaligo-NOAA The old instructions can
be found in older versions of ccpp-doc, such as at
https://ccpp-techdoc.readthedocs.io/en/v5.0.0/ParamSpecificOutput.html. You
can switch between versions of the doc in the lower right corner.
On Tue, Jul 27, 2021 at 11:09 AM ericaligo-NOAA ***@***.***>
wrote:
… Great. Thanks!
On 7/27/2021 1:03 PM, Samuel Trahan (NOAA contractor) wrote:
>
> Please file a ticket in the ccpp-doc issue page. The documentation you
> want still exists in an earlier version of the documentation
> repository (hash 3e9224c, I think) so the website would have to be
> updated... somehow... to display older versions.
>
> https://github.com/NCAR/ccpp-doc <https://github.com/NCAR/ccpp-doc>
>
> —
> You are receiving this because you were mentioned.
> Reply to this email directly, view it on GitHub
> <
#710 (reply in thread)>,
> or unsubscribe
> <
https://github.com/notifications/unsubscribe-auth/ALQ75MITLYZJQFP6DLJVX7DTZ3RE5ANCNFSM5BBCGXLA
>.
>
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#710 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AE7WQAS6QTVTAIY4CFPDRYDTZ3RZ7ANCNFSM5BBCGXLA>
.
|
Beta Was this translation helpful? Give feedback.
-
I would like to output the 3D non physics temperature tendency. I set ldiag3d=T and added dtend_select as follows:
&gfs_physics_nml
fhzero = 3.
ldiag3d = .true.
dtend_select = 'dtend_temp_nophys'
In diag_table, I have this:
"gfs_dyn", "dtend_temp_nophys", "dtend_temp_nophys", "fv3_history", "all", .false., "none", 2
Here's the error message I get:
859: forrtl: severe (19): invalid reference to variable in NAMELIST input, unit -5, file Internal Formatted NML Read, line -1, position 21
Beta Was this translation helpful? Give feedback.
All reactions