open boundary for future projection of CMIP6 downscaling

[fLy0516]

Hi,

Recently I used CMIP6 BCC results to do some future projection business. When the open boundary created by CMIP BCC ocean data, there is a blowing-up problem. However, changing the boundary file from SODA data, the model is able to run successfully. For future projection, maybe it's unreasonable to use SODA boundary directly, so how should the ocean boundary conditions be produced for future predictions to solve the above problem?

Thank you in advance.

[kate]

Well, I also hope to use CMIP for open boundaries in the future. Which of the models did you try and what software did you use to create the OBC files? Did the blow-up happen at the boundary and what did it look like?

[fLy0516]

I used ROMS3.5 and modified the program from CROCO ROMSTOOLS https://www.croco-ocean.org/ to create OBC file. It seems the blow-up did not happen at the boundary. My grid resolution is 0289x0379x0020, with open boundary at the south and the east, the blow-up message as shown below.

STEP Day HH:MM:SS KINETIC_ENRG POTEN_ENRG TOTAL_ENRG NET_VOLUME
C => (i,j,k) Cu Cv Cw Max Speed
500 3 11:20:00 2.083983E-01 1.850713E+04 1.850734E+04 9.947280E+15
(264,224,12) 2.399465E-02 1.084288E-01 9.278060E-01 7.455566E+00
510 3 13:00:00 2.102562E-01 1.850704E+04 1.850725E+04 9.947190E+15
(264,224,11) 2.189355E-02 1.000643E-01 9.042710E-01 7.454718E+00
520 3 14:40:00 2.124644E-01 1.850698E+04 1.850719E+04 9.947117E+15
(264,224,11) 1.840113E-02 9.620606E-02 8.687252E-01 7.591736E+00
530 3 16:20:00 2.147914E-01 1.850693E+04 1.850714E+04 9.947053E+15
(264,224,11) 1.759023E-02 9.428657E-02 8.223841E-01 7.698162E+00
540 3 18:00:00 2.172648E-01 1.850687E+04 1.850709E+04 9.947002E+15
(278,223,14) 3.532359E-01 1.476273E-01 9.399762E-01 7.843216E+00
550 3 19:40:00 2.343826E-01 1.850682E+04 1.850705E+04 9.946970E+15
(274,226,10) 2.634454E-01 1.375161E+00 1.826912E+01 5.392203E+01

Blowing-up: Saving latest model state into RESTART file

WRT_RST - wrote re-start fields (Index=1,1) into time record = 0000001

[kate]

The blow-up messages are not helpful. You need to look at the fields in the restart record that it saves.

[fLy0516]

zeta from restart file is extremely abnormal at open boundary ,

u/v are unreasonable throughout the region.

temp/salt are plausible.

figures as shown below.

[kate]

Salt looks like you have boundary weirdness, low values right at the edge. Zeta seems to be in balance with a strong boundary current which I'm not seeing in the u, v plots. Have you looked at the boundary files to see if the fields there look reasonable?

What values did you pick for LBC all around?

[fLy0516]

Yes, you are right. there are strong current speed at the open boundaries,and which course the zeta is large. current is shown as following.

The values for LBC, I guess, are these:

EAST_FSCHAPMAN Eastern edge, free-surface, Chapman condition.
EAST_M2FLATHER Eastern edge, 2D momentum, Flather condition.
EAST_M3NUDGING Eastern edge, 3D momentum, passive/active outflow/inflow.
EAST_M3RADIATION Eastern edge, 3D momentum, radiation condition.
EAST_TNUDGING Eastern edge, tracers, passive/active outflow/inflow.
EAST_TRADIATION Eastern edge, tracers, radiation condition.
SOUTH_FSCHAPMAN Southern edge, free-surface, Chapman condition.
SOUTH_M2FLATHER Southern edge, 2D momentum, Flather condition.
SOUTH_M3NUDGING Southern edge, 3D momentum, passive/active outflow/inflow.
SOUTH_M3RADIATION Southern edge, 3D momentum, radiation condition.
SOUTH_TNUDGING Southern edge, tracers, passive/active outflow/inflow.
SOUTH_TRADIATION Southern edge, tracers, radiation condition.

north and west are boundary wall.

[rtoste]

Are you using an initial condition also extracted from CMIP? If you are not using, you will have a abnormal gradient in zeta due to the different reference levels.

[fLy0516]

Yes, I'm using the initial condition from CMIP.

By comparing the u/v from CMIP and SODA, I found u/v from CMIP is about 0.5m/s smaller than soda,
Is it possible that the smaller current speed courses the blow-up?

[fLy0516]

According to rtoste's suggestion, I recreated the initial and boundary files from CMIP for my simulation, then it works well without blow-up problem.

[hpd14thu]

According to rtoste's suggestion, I recreated the initial and boundary files from CMIP for my simulation, then it works well without blow-up problem.

Good evening, I have the same question. I wonder that if your inition data and boundary data are from different database, which cause this problem?

Thank you!

[fLy0516]

I used the CIMP BCC data to create ini and bry files , which came from the same dataset.