Blowup debug

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
bperfect
Posts: 11
Joined: Fri Feb 19, 2016 4:05 pm
Location: University of Washington

Blowup debug

#1 Unread post by bperfect »

Is there a resource out there that lists some common knobs to turn to try to prevent a model blowup?

The ones that are apparent to me are:

timestep/Courant number
boundary conditions
viscosity coefficients
OBCFAC

Is there something else I am missing? I am running a 3D model with an idealized domain, and temp as my only tracer. The model runs fine in the absence of topography, but even a small hill produces nonphysical oscillations in the domain that eventually blow up.

My boundary conditions use the Flather/Chapman scheme in FSOBC and M2OBC to produce a constant flow entering from the western boundary. All other BC's are set to gradient.
Attachments
w_oscillations.png
u_oscillations.png
topography.png
topography.png (32.35 KiB) Viewed 1689 times

Post Reply