How the different arpegealadinalaroarome physical packages are coexisting

About This Presentation

Title:

How the different arpegealadinalaroarome physical packages are coexisting

Description:

Data-flow structures (90-100%) Dyn-phys interface (between 20 to ... The current interface routine mf_phys: the different physical packages are called by bloc ... – PowerPoint PPT presentation

Number of Views:33

Avg rating:3.0/5.0

Slides: 8

Provided by: Hel14

Category:

more less

Transcript and Presenter's Notes

Title: How the different arpegealadinalaroarome physical packages are coexisting

1
How the different (arpege/aladin/alaro/arome)
physical packages are co-existing

Short introduction to the 3 subsequent
presentations
Gwenaëlle Hello, Météo-France

2
Some generalities

Main part of the code is common
Setup (90)
Dynamics (100 )
Data-flow structures (90-100)
Dyn-phys interface (between 20 to 100 case
dependant)
Physic computations (between 0 to 100 case
dependant)
The diversity allowed in the setup is active in
the gridpoint computations

3
The setup
Level 0 CNT0
On/off Control high-level keys of the different
packages
su0yomA
SU0PHY
Physic detailed set-up
su0yomB
SUPHY
suphmf
suphmnh
suphec
4
The phys-gp computations, the drivers
GP-computations driver Gp_model
surfext
Arome
Arome
cpg
mf_phys
Arpege
Valid for
Aladin
cpg_pt
Alaro
cpg_dia
Arpege
Aladin
Alaro
5
The current interface routine mf_phys the
different physical packages are called by bloc
Arome
Arpege
Aladin
Alaro
mf_phys
aplpar
AAA
The diversity is inside
cptend Flux? tendencies
AAA
n -call to cputqy time evolution
AAA
apl_arome
A
n -calls to cputqy_arome
A
6
The data-flow the carriages are the same but the
passengers not as well as the road followed
Arpege
Arome
gmv/gfl
Aladin
Alaro
aplpar
Apl_arome
Flux
tends
cptend
cputqy_arome
tends
cputqy
Gmvt1/gflt1, Sl buf
7
Concluding remarks

Until now, the co-existence of the different
physical packages was with few difficulties to
handle
From dynamic side physics is seen as a
package so phys-dyn interaction is limited
As a consequence few degrees of freedom are
allowed (seq or //, bef or after dyn, gridpoint
or origin ?)
The physical packages are either quasi the same
(arp-ald), either with no intersection (arp-aro)
These last 3 points are less and less true
3D resolved physics will make the phys-dyn
interaction more complex
Then we will add more degrees of freedom
We are also going to add more and more physics
that we will also want to possibly mix