[RegCNET] 25 Km not permitted

[Graziano Giuliani]

On 10/08/2013 10:29 AM, abubakr salih wrote:
> Hi everyone,
> I tried to run RegCM4.3.5 in 25 Km resolution but it crashed due to hydrostatic
> limitation, though I thought the limit of the hydrostatic is on 10 Km. What it
> takes to run the model in 25 Km?
[...]

The stability of the dynamical solution depends also on the model topography and 
on the resolution of the driving global model and at least for regcm on the 
deformation due to map projection (the map factor).

1) Strong vertical topography gradient requires smaller time-steps. The 
Himalayas chain for one requires very small time-steps to allow the model to 
run. The usual ROT rule of "dt = 3*dx" works only with a smooth topography.

2) The downscaling usually can go as far as 1/5 of the original GCM resolution, 
to obtain higher resolutions you must plan to do a nesting, i.e. first run a 
coarse resolution and then use its ATM output as BC for the higher resolution.

3) Check that your domain keeps (look in the DOMAIN file) the map factors in the 
range 0.9-1.1 : higher/lower values especially on topography gradients are for 
sure source of stability problems.

4) Double check the input ICBC files for missing data/problems in the input 
conditions. The PreProc assumes the GCM data to be correct, but especially with 
binary dataset, a file corruption do not generate any warning/problem in the 3D 
interpolation engine, resulting in garbage input dataset.

Ciao!

-- 
Graziano Giuliani - Earth System Physics Section
The Abdus Salam International Centre for Theoretical Physics
Strada Costiera, 11 - I - 34151 Trieste Italy