Error in make whith analytic sea surface height actived

[Scarlett]

Hello modelers!!

I have a problem in make when I active analytic sea surface height, I get:

-------------------------------------------------------------------------
cd Build; /opt/pb/gcc-5.1.0/bin/gfortran -c -frepack-arrays -fopenmp -O3 -ffast-math -ffree-form -ffree-line-length-none mp_exchange.f90
cd Build; /opt/pb/gcc-5.1.0/bin/gfortran -c -frepack-arrays -fopenmp -O3 -ffast-math -ffree-form -ffree-line-length-none analytical.f90
analytical.f9032:

CLIMA(ng)%ssh(i,j)=???
1
Error: Invalid character in name at (1)
make: *** [Build/analytical.o] Error 1
-------------------------------------------------------------------------

I have already check analytical.f90 file in Ln 1100 and Col 32, I have:

!-----------------------------------------------------------------------
! Set sea surface height (meters).
!-----------------------------------------------------------------------
!
IF (LsshCLM(ng)) THEN
DO j=JstrT,JendT
DO i=IstrT,IendT
CLIMA(ng)%ssh(i,j)=???
END DO
END DO

Do I need establish a value in "CLIMA(ng)%ssh(i,j)=???"?
I hope you can help me, thanks in advance.

Mar.Mo.

[kate]

You should not be editing analytical.f90. It is created at compile time from analytical.F and all the ana_xxx.h files that get included there. You should edit ana_ssh.h which has that code you quote with the ???. This code was chosen to be something that doesn't compile, so that you know to go fix it to be correct for your case. Most of the ana_xxx.h files have various #ifdef MYDOMAIN sorts of sections, but none of the standard supported domains use ana_ssh.

[Scarlett]

Kate, so many thanks for your time and for your help. It works but I have a problem, in my history file, the salinity downs and temperature cools down as time passes (beginning by the borders), the superficial layers begin to cool and little by little the deepest ones.
Do you know why this happens? I have activated some analytical parameters such as:

#define ANA_BSFLUX /*Define for analytical bottom salt flux*/
#define ANA_BTFLUX /*Define for analytical bottom heat flux*/
#define ANA_DQDSST /*Define for an analytic surface heat flux sensitivy to SST*/
#define ANA_LRFLUX /*Define for analytic longwave radiation*/
#define ANA_SMFLUX /*Define for an analytic kinematic surface momentum stress*/
#define ANA_SRFLUX /*Define for an analytic kinematic surface shortwave radiation*/
#define ANA_SSH /*Define for an analytic sea surface height*/
#define ANA_SSS /*Define for an analytic sea surface salinity*/
#define ANA_SST /*Define for an analytic SST and ∂Q/∂SST*/
#define ANA_STFLUX /*Define for an analytic kinematic surface heat flux*/

Thanks in advance.
Mar.Mo.

[kate]

I would need to know more about your boundary conditions to understand what is happening. What boundary conditions are you using?

[Scarlett]

I added the lateral boundary conditions and the climatology files of the Hycom data, I have already verified this data and they are fine. The values of temperature and salinity are correct, but after the simulation I get something similar than I showed before.
I think I need to specify more for heat flux.

[lalvarez]

Hola Scarlett

What are your OBC's?
Are you using a sponge layer?
Once ago Kate suggested me to use RadNud and it worked goog for my domain.
Later on i'll send you my *.in as an example

Have a nice day

[Scarlett]

Hi Leo:
So many thanks for your help, I had Flather (2D momentum) for all state variables, I changed Flather for Radiation in temperature and salinity, and I increased the horizontal viscosity coefficient to 50 m2/s, and it works.
But do you suguest me to use RadNud? What's is the diference between RadNud and Flather?
Thanks in advance =)
Mar.Mo.

[kate]

Flather is only an option for 2d momentum. You need to pick Chapman for free surface and something like RadNud for the 3d variables. Flather and RadNud both require external values to work, which helps them behave better than something like pure Rad.

[Scarlett]

So many thanks Kate, I will try with it.
Mar.Mo.