Different boundary scenarios

[AMIRHOSSEIN]

I had a question about boundary conditions and according to the topic I asked earlier in the forum (ERA5_wind_problem):
viewtopic.php?t=5998
According to wilkin's tips, I realized that I would get better results if I used the Rad mode and also changed the visc value. But the problems persist. For example, I used Rad for the free surface, but after about 11 days of running the model, the results were not obtained correctly, and the same was true for the u and ubar values.

If I use the boundary as follows (in case we only have tide) we will have relatively better results(shown below the boundary conditions).

Code: Select all

   LBC(isFsur) ==   Clo     Cla     Cla     Clo         ! free-surface
   LBC(isUbar) ==   Clo     Fla     Fla     Clo         ! 2D U-momentum
   LBC(isVbar) ==   Clo     Fla     Fla     Clo         ! 2D V-momentum
   LBC(isUvel) ==   Clo     Fla     Fla     Clo         ! 3D U-momentum
   LBC(isVvel) ==   Clo     Fla     Fla     Clo         ! 3D V-momentum
   LBC(isMtke) ==   Clo     Gra     Gra     Clo         ! mixing TKE

   LBC(isTvar) ==   Clo     Clo     Clo     Clo \       ! temperature
                    Clo     Clo     Clo     Clo         ! salinity

free surface(zeta)

zeta.JPG (71.61 KiB) Viewed 27235 times

ubar

ubar.JPG (73.2 KiB) Viewed 27235 times

Based on the results, I decided to give the ERA5 information to the model and activate BULK_FLUXE in the file header to see the changes in the zeta, ubar, and u parameters. Finally, with the assumptions of Clamped mode for free surface and Flather for all others modes, I encountered a maxspeed error on the first day of model operation. However, after this performance, I again considered all boundary modes as Rad, but again, I did not get the desired results.

ubar

ubar(wind-all-Rad).JPG (72.04 KiB) Viewed 27235 times

free surface(zeta)

free surface(wind-all-Rad).JPG (75.3 KiB) Viewed 27235 times

tide-Era5-allRad.log

logfile

(855.17 KiB) Downloaded 412 times

Now I'm a little confused about what part of the model to focus on and think about improving it! Is the problem from the boundary file or in choosing the type of boundary?
According to the above description, if I use Rad, the boundary file will no longer be read by the model. Does this mean I do not need a border file? (however, I have only velocity data from TMD in my boundary.nc file)

Thank you for taking the time to read this issue.

[kate]

For what it's worth, this is what I'm using:

LBC(isFsur) == Che Che Che Che ! free-surface
LBC(isUbar) == Shc Shc Shc Shc ! 2D U-momentum
LBC(isVbar) == Shc Shc Shc Shc ! 2D V-momentum
LBC(isUvel) == RadNud RadNud RadNud RadNud ! 3D U-momentum
LBC(isVvel) == RadNud RadNud RadNud RadNud ! 3D V-momentum
LBC(isMtke) == Clo Clo Clo Clo ! mixing TKE

LBC(isTvar) == RadNud RadNud RadNud RadNud \ ! temperature
RadNud RadNud RadNud RadNud \ ! salinity

Not all options are valid for all variables.

[AMIRHOSSEIN]

Hi kate,
thanks for your instant response. As you said, I changed the boundary conditions, but nothing changed.
I am waiting for your guidance.

kate.log

my log file

(1.53 MiB) Downloaded 495 times

thanks

[jcwarner]

on this page, top figure, what is 'wrong' with the zeta?

[AMIRHOSSEIN]

Hi jcwarner, I am very thankful you are considering my problem.
I apologize for being late. Something was wrong with my laptop.
The problem with zeta is that after 11 days its amounts are very small. Before running the model in 3D, I ran it in 2D for 30 days
and zeta changes were between 2 to 2.3 meters(with the only tide as input). After that, I ran the in 3D(like in 2D) and zeta changes
were between 1.5 to 1.8 meters. finally, I defined BULK_FLUX and used Era5 data. As you see in the top figure, The zeta value should be
much higher than these values. shouldn't be?!

Thanks

[jcwarner]

think you are changing too many things here.
ERA winds should not affect the tides. and dont rely on ncview to give you the max/min. click at a point and get a time series.
try this:
You have this:
Variable Grid West Edge South Edge East Edge North Edge
--------- ---- ---------- ---------- ---------- ----------
zeta 1 Closed Radiation Radiation Closed
ubar 1 Closed Radiation Radiation Closed
vbar 1 Closed Radiation Radiation Closed
u 1 Closed Radiation Radiation Closed
v 1 Closed Radiation Radiation Closed
temp 1 Closed Closed Closed Closed
salt 1 Closed Closed Closed Closed
tke 1 Closed Radiation Radiation Closed

You need to use
zeta to be Chapman for all the open sides.
ubar and vbar to be Flather for all the open sides.
temp/salt should be nudged to something, or just set to Rad to start with.

set ndtfast to be 20

and give that a try.

[asik759495]

Asik

[AMIRHOSSEIN]

Thanks for your reply jcwarner,
Regarding your last sentence about salinity and temperature, do you mean that I should take salinity and temperature data
from databases such as Hycom or something else and enter it in the model as an input file?

[AMIRHOSSEIN]

think you are changing too many things here.
ERA winds should not affect the tides. and dont rely on ncview to give you the max/min. click at a point and get a time series.
try this:
You have this:
Variable Grid West Edge South Edge East Edge North Edge
--------- ---- ---------- ---------- ---------- ----------
zeta 1 Closed Radiation Radiation Closed
ubar 1 Closed Radiation Radiation Closed
vbar 1 Closed Radiation Radiation Closed
u 1 Closed Radiation Radiation Closed
v 1 Closed Radiation Radiation Closed
temp 1 Closed Closed Closed Closed
salt 1 Closed Closed Closed Closed
tke 1 Closed Radiation Radiation Closed

You need to use
zeta to be Chapman for all the open sides.
ubar and vbar to be Flather for all the open sides.
temp/salt should be nudged to something, or just set to Rad to start with.

set ndtfast to be 20

and give that a try.

Hi jcwarner,
After a relatively long time on the third day of the model, I again encountered the maxspeed error.

Code: Select all

     18378 2013-01-03 03:03:00.00  2.505877E-02  1.246039E+04  1.246042E+04  2.406096E+14
                     (020,283,10)  4.066740E-03  1.617839E-02  9.361253E-02  7.348638E+00
     18379 2013-01-03 03:03:10.00  2.506269E-02  1.246040E+04  1.246042E+04  2.406095E+14
                     (039,301,10)  1.173852E-03  4.500941E-02  2.149429E-01  1.700259E+01
     18380 2013-01-03 03:03:20.00  2.506633E-02  1.246040E+04  1.246043E+04  2.406094E+14
                     (039,301,10)  1.315355E-03  1.185552E-01  7.354894E-01  4.650887E+01
 Found Error: 01   Line: 321      Source: ROMS/Nonlinear/main3d.F
 Found Error: 01   Line: 298      Source: ROMS/Drivers/nl_roms.h, ROMS_run

 Blowing-up: Saving latest model state into  RESTART file
     REASON: MaxSpeed =  4.7E+01

and here are zeta, ubar and u from ncview:

zetafinal.JPG (68.35 KiB) Viewed 26766 times

Time series near the Strait of Hormuz

Time series near the Strait of Hormuz

[jcwarner]

do you have
#define CURVGRID
that might help with some of the advection.
take a look at the winds in that ncviewer. how strong are they on the west side?

-j

[AMIRHOSSEIN]

do you have
#define CURVGRID
that might help with some of the advection.
take a look at the winds in that ncviewer. how strong are they on the west side?

-j

thank you for your reply jcwarner,
As you said for defining CURVGRID in the header file, the angle in my grid is zero, and is it still advisable to enable this option?
and u10 time series from my Era5 data in the west is shown here:

u10-domain

u10.JPG (75.82 KiB) Viewed 26685 times

time-series-Uwind

[AMIRHOSSEIN]

Hi jcwarner,
As you said, I also defined CURVGRID and here is the result:

Code: Select all

      18284 2013-01-03 02:47:20.00  2.468579E-02  1.246003E+04  1.246006E+04  2.406199E+14
                     (020,283,10)  3.895822E-03  1.595643E-02  7.901521E-02  8.782318E+00
     18285 2013-01-03 02:47:30.00  2.469043E-02  1.246003E+04  1.246006E+04  2.406198E+14
                     (039,301,10)  1.580910E-03  2.909716E-02  2.835400E-01  1.324834E+01
     18286 2013-01-03 02:47:40.00  2.469398E-02  1.246004E+04  1.246006E+04  2.406196E+14
                     (039,301,10)  1.278797E-02  8.649580E-03  4.267355E-01  1.074167E+01
 Found Error: 01   Line: 321      Source: ROMS/Nonlinear/main3d.F
 Found Error: 01   Line: 298      Source: ROMS/Drivers/nl_roms.h, ROMS_run

 Blowing-up: Saving latest model state into  RESTART file
     REASON: MaxDensity =  2.7E+03

Thanks

[jcwarner]

I am nots really sure what all to suggest here.
A few images back , you had zeta on the west side at -8m.
Can you figure out why that happened? Even a super strong wind would have trouble to do that.
I thought the winds might not be rotated to the grid correctly.

Suggest you set roms to write out the winds to its own netcdf output, and compare the winds in the roms his file to the winds in the ERA forcing file.

[Kosa]

Are you using WET_DRY option? It's my guess since -10 m zeta didn't cause a blowup. If so, I would try without WET_DRY unless you actually need it. And then you will need to set a minimum depth so that no grid points become fully dry at low tide. So e.g. set all depths to be at least 3 m.

[AMIRHOSSEIN]

Thanks, dear jcwarner and Kosa for your tips. I'll do them and inform the results.

[AMIRHOSSEIN]

Hi everybody,
As jcwarner said, it was for the Era data, and it was because of the pressure that displayed 100 times larger in the ROMS output.

thanks a lot