How do I solve the simulation speed anomaly?

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
zhaolv
Posts: 39
Joined: Thu Nov 11, 2021 3:56 pm
Location: Changsha University of Science and Technology

How do I solve the simulation speed anomaly?

#1 Unread post by zhaolv »

I was running roms and found that the model crashed at the fourth hour and prompted me with excessive speed at a certain point, I checked the rst file and found that there was indeed an exception, how do I fix this?
I have tried changing the model grid, the time step, but nothing has fixed the problem. Please can you help me.
Below is a screenshot of my logs file and rst results.

Code: Select all

                     (015,032,16)  8.993006E-04  1.060675E-03  2.194779E-01  2.871628E+00
      1127 2018-09-15 04:41:45.00  1.049570E-01  1.322309E+04  1.322319E+04  4.267362E+14  02
                     (015,032,16)  9.087804E-04  1.077216E-03  2.204331E-01  5.698030E+00
      FINE2COARSE - exchanging data between grids: dg = 02 and rg = 01  at cr = 02
       564 2018-09-15 04:42:00.00  1.207169E-01  1.518843E+04  1.518855E+04  1.468960E+15  01
                      (043,41,16)  5.148341E-04  1.923969E-03  1.325345E-01  4.506861E+01
 DIAG speed trouble    45.068605085211125     
 DIAG speed ijk    63.723548390533701               30          27           1           1
 Found Error: 01   Line: 341      Source: ROMS/Nonlinear/main3d.F
 Found Error: 01   Line: 331      Source: ROMS/Drivers/nl_ocean.h, ROMS_run

 Blowing-up: Saving latest model state into  RESTART file
     REASON: MaxSpeed =  4.5E+01

      WRT_RST     - wrote re-start    fields (Index=1,1) in record = 3                     01

 Blowing-up: Saving latest model state into  RESTART file
     REASON: MaxSpeed =  4.5E+01

      WRT_RST     - wrote re-start    fields (Index=1,2) in record = 3                     02
rst.png
rst.png (25.69 KiB) Viewed 3211 times
Attachments
roms.txt
(440.51 KiB) Downloaded 120 times

jcwarner
Posts: 1172
Joined: Wed Dec 31, 2003 6:16 pm
Location: USGS, USA

Re: How do I solve the simulation speed anomaly?

#2 Unread post by jcwarner »

what is the bathy in that corner?
i see you have
2 Vtransform S-coordinate transformation equation.
4 Vstretching S-coordinate stretching function.
5.0000E+00 theta_s S-coordinate surface control parameter.
4.0000E-01 theta_b S-coordinate bottom control parameter.
50.000 Tcline S-coordinate surface/bottom layer width (m) used
in vertical coordinate stretching.

not sure what that looks like in the vertical, but it might be tight vertical spacing in the shallow locations.

zhaolv
Posts: 39
Joined: Thu Nov 11, 2021 3:56 pm
Location: Changsha University of Science and Technology

Re: How do I solve the simulation speed anomaly?

#3 Unread post by zhaolv »

jcwarner wrote: Fri Apr 22, 2022 11:54 am what is the bathy in that corner?
i see you have
2 Vtransform S-coordinate transformation equation.
4 Vstretching S-coordinate stretching function.
5.0000E+00 theta_s S-coordinate surface control parameter.
4.0000E-01 theta_b S-coordinate bottom control parameter.
50.000 Tcline S-coordinate surface/bottom layer width (m) used
in vertical coordinate stretching.

not sure what that looks like in the vertical, but it might be tight vertical spacing in the shallow locations.
I referred to other solutions to the problem and turned on the WET_DRY switch in the cpp and things got somewhat better.
But after I expanded the grid range, I ran it again and found some problems with the northern boundary.
rst_1.png
rst_1.png (47.19 KiB) Viewed 3071 times

jcwarner
Posts: 1172
Joined: Wed Dec 31, 2003 6:16 pm
Location: USGS, USA

Re: How do I solve the simulation speed anomaly?

#4 Unread post by jcwarner »

ok. if you min depth = 5 meters, and then wet/dry helped, i suggest you look at the zeta values to make sure the tides are ok. what is the tide range in this area?

for the next problem, that is probably a density issue. closely look at what is being put along that boundary for temp salt u v etc.

you need to dig into the output.
-j

Post Reply