| Version 3 (modified by , 4 years ago) (diff) |
|---|
20th January 2022
Duration: 1h10mn
Participants: Jiandong, Lucas, Siteng, Romain, Christophe, Ehouarn, Margaux, Antoine, Thomas
Aymeric, Sandrine
News:
- Be ready to contribute in the coming weeks to the "Atmosphere planétaires" project
application to obtain HPC computing hours for 01/05/2022-30/04/2023 (Occigen
will shut down and we'll move to Irene-Rome)
Recent commits: https://trac.lmd.jussieu.fr/Planeto/log
==============
- r2605-2617: OpenMP directives and code adaptation to run with OpenMP enabled
Recent tickets: https://trac.lmd.jussieu.fr/Planeto/report/1
==============
- None
- ticket about 32um emissivity; Christophe will add more information
- also the ticket about initracer to add some information in
Hot topics:
==========
* OpenMP:(Romain)
- OpenMP, tests on LMD computers OK; some issues still on Occigen
Will make a presentation on using/coding with OpenMP next week
* Improving the water cycles (Antoine)
- water non-conservation issue: tests between beginning/end of physics show that
global planet-wide value changes within the dynamics.
to check: is the total mass of atmosphere conserved from 1 phys step to the next?
* Non-orographic GW with the thermosphere setup (Jiandong)
- Has found more tuning parameters to play with which "solve" the polar warming and
can affect the phase of the thermal tides (by hanging where momentum is deposited)
Jiandong:
- looking also at the orographic GW code (to check connections with non-oro GW).
Cleaning/Upgrading the code to F90 and documenting it!
Ehouarn:
- made new "clim", "cold", "warm" dust scenarios using the latest MY24-MY35 krieged dust maps
there are some not that minor differences with respect to the old (MCD5.*) ones; not that
straightforward to quantify as "old" dust scenarios are on 72x36 grids, whereas new ones are
on 120x60. It will probably make sense to have both "old" and "new" cold/clim/warm scenarios
in the GCM.
Margaux:
- submitted the HDO paper this morning
Lucas:
- submitted his paper on Insight pressure
- working with Romain on the PEM. Tests at low obliquity with atmopheric collapse
(which is not as fast as expected); comparing with E. Dupont's internship results with nogcm
- tests with subslopes look good; implementing water ice in preparation
Siteng:
- update on observations (TIRVIM) and comparisons with v5.3 GCM
- difference of phase between obs and model of ~1h (lag of the GCM wrt obs)
Thomas:
- slope winds model in MCD6 : comparisons with the mesoscale model from which a GCM-like
mean field has been subtracted; has devised a more general minimizing scheme to
evaluate best fit parameters. Will use it on comparisons with data at mesoscale levels
from near-surface to ~2km.
13th January 2022
Duration: 1h
Participants: Margaux, Ehouarn, Jiandong, Christophe, Romain, Thomas, Antoine, Sandrine, Antony, Lucas
Francois, Siteng
News:
- Deadline for GENCI application 14 Feb.
- Mike Smith and Mike Wolff will come to LMD 21-25th February,
- Mike Wolff will also stay at LMD in April and June.
- Open question to all: would you prefer to have the POULDP Mars meetings start at noon or 11am?
Recent commits: https://trac.lmd.jussieu.fr/Planeto/log
==============
- r2604: Stop with a clear error message if "traceur.def" file is missing
Recent tickets: https://trac.lmd.jussieu.fr/Planeto/report/1
==============
- CLOSED: Tickets which could have been closed earlier because issue was solved
#3, #4, #12, #13, #20, #21, #40, #59, #73, #74
Hot topics:
==========
* Water conservation (Antoine, Joseph)
- investigated the water conservation over 10 year; still a small conservation issue ~2%/year
compared with Joseph's simulations => same results : ~2% loss / year regardless of GCM53 or GCM6 dust parametrizations
also has the problem with 54 layers
- will check if the problem indeed comes from the dynamics and add a fix
* OpenMP:(Romain)
- maybe there is no MPI problem...
- will integrate the OpenMP changes step by step.
* Non-orographic GW with the thermosphere setup (Jiandong)
- investigating the ranges of impact of the parametrization: never gets the same phase of the thermal tide
as MCS. Notes that polar regions differ between GCM and MCS. => insufficient polar warming in the GCM
* Improving the dust cycle (Antoine)
- To be continued...
Ehouarn:
- testing Irene-Rome with "references setups" to check how much slower/faster it is than Occigen
to be confirmed, but it seems a little faster
(a big surprise to me as I've been told that LMDZ6, the Earth GCM is slower when run on Irene-Rome...)
- will make new "clim"/"cold"/"warm" scenarios using MY24-MY35 dust scenarios ASAP
Sandrine:
- resuming work on paper on thermal tides from TIRVIM data.
interested in checking out the results with the new non-oro GW once it is tuned.
Siteng:
- Update on temperature differences between TIRVIM & GCM with time of day
6th January 2022
Duration: 1h15mn
Participants: Tanguy, Margaux, Siteng, Aymeric, Antoine, Christophe, Ehouarn, Francois, Romain, Thomas
Jiandong, Lucas
News:
- Next week we will hold the POULDP Mars meeting earlier and start at 11am (targeting to end at noon
because Francois, Ehouarn, Aymeric won't be available after).
Recent commits:https://trac.lmd.jussieu.fr/Planeto/log
==============
- r2596: missing OpenMP THREADPRIVATE for Non-orographic GW scheme
- r2598: updated arch files for the Jean-Zay supercomputer
- r2599-2602: clean up of co2condens_mod and bug fix in physiq for
the scavenging by CO2 ice (when without CO2 ice clouds)
Recent tickets: https://trac.lmd.jussieu.fr/Planeto/report/1
==============
-TODO: clean up the tickets (some are old and obsolete but not closed)
Ehouarn, Antoine and whoever else wants to join, will start this afternoon at the Bocal.
Hot topics:
==========
* Water conservation (Antoine, Joseph)
- 10 year long runs with scavenging bug fixed have been made. No apparent (big) conservation problem.
Further analysis required to better tune the water cycle compared to TES
* Improving the dust cycle (Antoine)
- To be continued...
* OpenMP:(Romain)
- all seems ok now. To be commited.
will present an overview to all at a future Pouldp
* Non-orographic GW with the thermosphere setup (Jiandong)
- tests with MY29 yearly runs with various parameters, some fails because of huge winds ~700m/s at topmost layers,
which can also happen without the GW parametrization.
- comparisons with MCS temperature are ongoing
Thomas:
- found a bug in the MCD (VCD also) in the variability (RMS) of the density in the topmost layer, has
a correction to implement.
- MCD web interface for full version requests ready. New web interface as well. All ready to be put online
in parallel to the current one for tests from the team.
- slope winds: extracting "slope winds" from mesocale simulations by substracting large scale GCM-component.
Margaux:
- working on the HDO articles which will soon be submitted
Lucas:
- working on the comments by co-authors to his paper; will submit next week.
- still working on implementing the sub-slopes
Siteng:
- looks at the thermal response variations with dust content (simple homogeneous same dust everywhere case),
varying dust top altitude and tau.
