Hello,
I am running CROCO from an ini.nc file and the model becomes unstable immediately at timestep 1.
Model output:
STEP time[DAYS] KINETIC_ENRG POTEN_ENRG TOTAL_ENRG NET_VOLUME trd
0 2922.00000 1.937450459E-03 4.4991216E+01 4.4993153E+01 3.8496160E+14 0
1 2922.00069 NaN NaN NaN NaN 0
I have already checked the bry file and the ini file and they do not show any visible NaNs.
The only suspicious thing I can identify is when I plot croco_his.nc: I can see anomalous vertical velocity (W) near the northern and southern boundaries, while the other variables seem to be fine.
Has anyone experienced a similar issue or knows what could generate these anomalous W values at timestep 1?
Any suggestions would be greatly appreciated.
croco_Y2008M01.txt (17.8 KB)
croco_Y2008M01_inter.txt (10.5 KB)
I reduced the dt by half, but the error still persists.
Increase the at may be keep 150 sec or 300 sec
Did you get your issue solved?
I did change the dt to different values but the NaNs on the timestep 1 still continue
Hi,
ncdump -h croco_ini.nc | grep -E “hbl|hbbl|temp|salt|zeta|float|double”
share output.
and try gls mixing. see PGE still high or not.
Thank you.
ncdump -h croco_ini.nc | egrep “hbl|hbbl|temp|salt|zeta|float|double”
double s_rho(s_rho) ;
double Cs_rho(s_rho) ;
double Cs_w(s_w) ;
double theta_s(one) ;
double theta_b(one) ;
double Tcline(one) ;
double hc(one) ;
double tstart(one) ;
double tend(one) ;
double ocean_time(time) ;
double scrum_time(time) ;
double u(time, s_rho, eta_rho, xi_u) ;
double v(time, s_rho, eta_v, xi_rho) ;
double ubar(time, eta_rho, xi_u) ;
double vbar(time, eta_v, xi_rho) ;
double zeta(time, eta_rho, xi_rho) ;
zeta:long_name = “initial sea surface height” ;
zeta:units = “meters” ;
double temp(time, s_rho, eta_rho, xi_rho) ;
temp:long_name = “initial potential temperature” ;
temp:units = “Celsius” ;
double salt(time, s_rho, eta_rho, xi_rho) ;
salt:long_name = “initial salinity” ;
salt:units = “psu” ;
I’m going to try gls mixing and report back the outcome.
Cheers
I tried to enable GLS mixing by following the tutorial the croco documentation, but got an error
initarrays.f(1285): error #6410: This name has not been declared as an array or a function. [TRB]
trb(i,j,k,nstp,itke)=1.D-12
------------^
initarrays.f(1286): error #6410: This name has not been declared as an array or a function. [TRB]
trb(i,j,k,nstp,igls)=1.D-12
------------^
initarrays.f(1287): error #6410: This name has not been declared as an array or a function. [LSCALE]
Lscale(i,j,k)=init
------------^
compilation aborted for initarrays.f (code 1)
gmake: *** [Makedefs:15: init_arrays.o] Error 1
I did manage to find the cause of the error. I had undefined a key that was essential.
I would like to thank everyone for their help
Hi,
which key you undef . you need to choose GLS different methods. default is GLS_GEN. GLS_KOMEGA, GLS_KEPSILON. choose based on your application (cppdefs_dev.h or cppdefs.h)
Thank you.
Best regards,
Ultimately, I kept the default mixing for my run. The only changes I had to make were to define OBC_M2CHARACT and undefine OBC_M2ORLANSKI.
But thanks for mentioning the different GLS mixing!
Cheers
OK.
Now you have :
define OBC_M3ORLANSKI
define OBC_TORLANSKI
/* add this based your BC (change) */
define OBC_COM_WEST
define OBC_COM_EAST
define OBC_COM_SOUTH
define OBC_COM_NORTH
there also have problem in BC (remembering)
i did it before (wrong velocity in BC case)
i need to check that file where i store in my hpc.
i will share with you and let you know this week.
Thank you again.
Now, it is like this and running without any problem.
I did also define the OBC..west and everything
# define OBC_M2CHARACT
# undef OBC_M2ORLANSKI
# define OBC_M3ORLANSKI
# define OBC_TORLANSKI
# undef OBC_M2SPECIFIED
# undef OBC_M3SPECIFIED
# undef OBC_TSPECIFIED
Thank you
Perfect.
OBC_COM_M2CHARACT some way Chapman.
OBC_COM_M2CHARACT0 some way inflow-switch BC
(grep -r M2CHARACT0)
to use it you need to add in set_obc_definitions.h
for mapping. currently its dead code in CROCO:
# ifdef OBC_M2CHARACT0
# define OBC_COM_M2CHARACT0
# endif
i will share correct BC file.
please give me this week. to much work i have this week.
meet you next week. sorry.
Best regards.
