Hi,
I am running a nested version of ROMS around the faroe shelf and my nested domain keeps blowing up, with out any obvious reason. I have tried to reduce the Haney number. but then it still blows up. I would like keep at least 35 vertical layers, and have 400 m horizontal resolution. The area has some steep bathymetry.
I have run a 2kmx2km domain that runs fine, which I use as boundary forcing, if I reduce the time step to 0.01 seconds it runs for 17 timesteps.
The errors I get are:
Found Error: 01 Line: 314 Source: ROMS/Nonlinear/main3d.F
Found Error: 01 Line: 302 Source: ROMS/Drivers/nl_ocean.h, ROMS_run
ROMS/TOMS - Output NetCDF summary for Grid 01:
number of time records written in HISTORY file = 4
number of time records written in STATION file = 1
Analytical header files used:
ROMS/Functionals/ana_btflux.h
ROMS/Functionals/ana_srflux.h
Found Error: 01 Line: 361 Source: ROMS/Utility/close_io.F, close_out
ROMS/TOMS - Blows up ................ exit_flag: 1
Basin information for Grid 01:
Maximum grid stiffness ratios: rx0 = 8.964749E-01 (Beckmann and Haidvogel)
rx1 = 2.926206E+01 (Haney)
Initial domain volumes: TotVolume = 1.8212118841E+12 m3
MinCellVol = 2.2047829491E+04 m3
MaxCellVol = 8.9615248835E+06 m3
Max/Min = 4.0645837211E+02
Blow up in nested version
Blow up in nested version
- Attachments
-
-
out.dat- (122.2 KiB) Downloaded 188 times
Re: Blow up in nested version
Hi. Here are some questions to consider.
1) Does the nested grid run in a stand-alone configuration (i.e., not in a nested configuration)? In other words, is this an issue with the nesting or the domain?
2) Is the value of rx0 given for the nested domain or for the outer domain? I'll assume it is for the nest.
3) The numbers you quote for rx0 and rx1 are on the high side in my opinion, though there are applications that run with numbers like that. You might examine if these (delta h/h) is large in many areas around the island or just in a few. You might try smoothing the bathymetry considerable to see if it runs.
4) What is the maximum Courant Number? If it is close to 1, a locally fast current may cause the CFL condition to be tripped. You might try reducing the time-step by a factor of 10 and see if it runs.
5) Finally, have you looked at the location where the blowup happens? This requires saving fields around the time when the blowup occurs. It often requires doing a restart to get close to the blowup and then saving fields very frequently. Often seeing where and how a blowup begins can help to find a solution.
I hope these suggestions help. Good luck!
Chris
1) Does the nested grid run in a stand-alone configuration (i.e., not in a nested configuration)? In other words, is this an issue with the nesting or the domain?
2) Is the value of rx0 given for the nested domain or for the outer domain? I'll assume it is for the nest.
3) The numbers you quote for rx0 and rx1 are on the high side in my opinion, though there are applications that run with numbers like that. You might examine if these (delta h/h) is large in many areas around the island or just in a few. You might try smoothing the bathymetry considerable to see if it runs.
4) What is the maximum Courant Number? If it is close to 1, a locally fast current may cause the CFL condition to be tripped. You might try reducing the time-step by a factor of 10 and see if it runs.
5) Finally, have you looked at the location where the blowup happens? This requires saving fields around the time when the blowup occurs. It often requires doing a restart to get close to the blowup and then saving fields very frequently. Often seeing where and how a blowup begins can help to find a solution.
I hope these suggestions help. Good luck!
Chris