[RegCNET] CFL>1 violation error
Chao Tang
chao.tang at univ-reunion.fr
Wed Sep 21 18:49:37 CEST 2016
Dear all,
I got CFL>1 euro bellow, when running RegCM-4.5.0:
===================================================
Initial date of the global simulation: 2004-10-02 00:00:00 UTC
Initial date of this run : 2004-10-02 00:00:00 UTC
Final date of this run : 2005-07-01 00:00:00 UTC
Total simulation lenght : 6528 hours
Timestep in seconds = 2.000000
Reading Domain file : input/LaReunion_DOMAIN000.nc
Setting IPCC scenario to A1B
…
Boundary Pameterizations
Nudge value high range : 3.000000
Nudge value medium range : 2.000000
Nudge value low range : 1.000000
Model Timestep Pameterizations
time step for dynamical model in seconds : 2.000000
time step for surface model in seconds : 600.000000
time step for cumulus model in seconds : 300.000000
time step for radiation model in minutes : 30.000000
time step for emission model in hours : 6.000000
Domain grid parameters:
Map Projection : LAMCON
Dot Grid Full Extent : 260x 160x 18
Model Top Pressure : 5.000000 cb
Model Grid Spacing : 3.000000 km
Grid Center Latitude : -20.800000 deg
Grid Center longitude : 56.300000 deg
True Latitude 1 : 30.000000 deg
True Latitude 2 : 60.000000 deg
…
$$ no. of points with active convection = 7557
SEARCH BC data for 2004101400, ktau = 507600
READY BC from 2004101318 to 2004101400
ATM variables written at 2004-10-13 18:00:00 UTC
SRF variables written at 2004-10-13 18:00:00 UTC
RAD variables written at 2004-10-13 18:00:00 UTC
$$$ 2004-10-13 18:00:00 UTC, ktau = 507600
$$$ max value of CFL = 0.26060E-01
$$$ no. of points with active convection = 5389
CFL>1: CFL = 1.0019 W = -187.4752 I = 49 J = 91 K = 14
CFL>1: CFL = 1.0242 W = -144.4658 I = 49 J = 91 K = 15
-------------- FATAL CALLED ---------------
Fatal in file: mod_sound.F90 at line: 686
CFL violation
-------------------------------------------
Abort called by computing node 34 at 2016-09-21 14:37:20.171 +0200
—————————————————————————————————————
As shown above:
1. ds=3km, dt=2s, domain=260*160*18 grid;
2. the simulation starts from 2004-10-02, and stopped around 2004-10-13.
Questions:
1, For CFL in RegCM, I/J are number of grid, K is the vertical level ?
2, what the meaning of W ?
3, How this 3D CFL calculated in RegCM-4.5.0 ?
4, Any idea to fix this error ?
Thanks a lot in advance .
Cheers,
Chao TANG
University of La Reunion, France.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20160921/723ce7e4/attachment.html>
More information about the RegCNET
mailing list