open TELEMAC-MASCARET The mathematically superior suite of solvers

I saw the keyword STOCHASTIC DIFFUSION MODEL in TELEMAC2D user manual, but I cannot see this keyword in TELEMAC3D. Does it mean that there is a stochastic diffusion model in TELEMAC2D but not TELEMAC3D? Or the TELEMAC3D manual just miss this part?

If TELEMAC3D does not have a stochastic diffusion model, any suggestions as to how can I do particle tracking with stochastic diffusion model?

Hello Arthur,

Indeed, no such keyword in TELEMAC-3D because this feature has not existed yet. Talking quickly to people who khow this feature in 2D, the issue is how to define it over the vertical in 3D.

I cannot help you more.

Chi-Tuan

Thank you. Do you have any experience in using a Opensoure particle tracking software OpenDrift? It seems to be good but I am not sure if it is compatible with the output from TELEMAC3D.

The link of the software is attached below:
opendrift.github.io/index.html

Hello Arthur,

Sorry, but I did not know this tool.
Using what is written in hermes module and Python scripts may link TELEMAC-3D with OpenDrift, you can try.

Best regards,

Chi-Tuan

Could you please explain more about the "hermes module" and "Python script"?

Hermes is a module that gives you generic features to handle different file formats available for the TELEMAC system (e.g.: SERAFIN, MED). You can read the developer_guide (chapter 9). The source codes are located in $HOMETEL/sources/utils/hermes

Python scripts are located in $HOMETEL/scripts/python3
You can read the notebooks to see what is available ($HOMETEL/notebook). E.g.: run jupyter-notebook index.ipynb to read the index.

Hope this helps,

Chi-Tuan

Hi,

I realise that stochastic diffusion is not implemented in Telemac3D but I still would like to see the influence of this in the horizontal plane in a 3D study.

I have copied streamline.f into my user_fortran directory and modified the SCHAR_41 subroutine by adding:

///////////////////////////////////////////////////////////
DZ(IPLOT) = ((W(I1,IET )*SHP(1,IPLOT)
& + W(I2,IET )*SHP(2,IPLOT)
& + W(I3,IET )*SHP(3,IPLOT))*(1.D0-SHZ(IPLOT))
& +(W(I1,IET+1)*SHP(1,IPLOT)
& + W(I2,IET+1)*SHP(2,IPLOT)
& + W(I3,IET+1)*SHP(3,IPLOT))*SHZ(IPLOT) ) * PAS
ENDIF
!
! STOCHASTIC DIFFUSION
!
! IF(STOCHA.EQ.1) THEN
! COMPUTING LOCAL VISCOSITY
! A=MAX(VISCVI%ADR(1)%P%R(I1)*SHP(1,IPLOT)
! & +VISCVI%ADR(1)%P%R(I2)*SHP(2,IPLOT)
! & +VISCVI%ADR(1)%P%R(I3)*SHP(3,IPLOT),0.D0)
! B=MAX(VISCVI%ADR(2)%P%R(I1)*SHP(1,IPLOT)
! & +VISCVI%ADR(2)%P%R(I2)*SHP(2,IPLOT)
! & +VISCVI%ADR(2)%P%R(I3)*SHP(3,IPLOT),0.D0)
! DISPLACEMENT DUE TO RANDOM DIFFUSION
A=0.6D0
B=0.6D0
DEUXPI=2.D0*ACOS(-1.D0)
CALL RANDOM_NUMBER(RAND1)
CALL RANDOM_NUMBER(RAND2)
C=SQRT(-2.D0*LOG(RAND1))
D=C*COS(DEUXPI*RAND2)
E=C*SIN(DEUXPI*RAND2)
DIFF_X=D*SQRT(1.D0*A/1.D0)
DIFF_Y=E*SQRT(1.D0*B/1.D0)
DX(IPLOT) = DX(IPLOT) + DIFF_X*SQRT(ABS(PAS))
DY(IPLOT) = DY(IPLOT) + DIFF_Y*SQRT(ABS(PAS))
! ENDIF

WRITE (LU,*) 'HERE!:'

XP = XPLOT(IPLOT) + DX(IPLOT)
YP = YPLOT(IPLOT) + DY(IPLOT)
ZP = ZPLOT(IPLOT) + DZ(IPLOT)
//////////////////////////////////////////////////////

Adding the WRITE statement was to check that the code was being accessed but it doesn't appear to be reaching this line or using the SCHAR_41 subroutine when running my case.

Is this the correct way to do this or am I missing something?

Thanks for any help or suggestions,

Tom

Hello Tom,

Have you run your computation in parallel mode? If yes, have you looked at every PE***.LOG output files in the temporary folder? Your writing may be in subdomains other than the master.

Hope this helps,

Chi-Tuan

Hi Chi-Tuan,

Thanks for your reply. I was running in serial mode.

Best,

Tom

Tom,

Can you send your steering file + modified STREAMLINES subroutine please?
+ give the release you use (v8p1 I hope).

As you talk about SCHAR_41 subroutine, I do not know if you mean SCHAR41 (without _), SCHAR41_STO or another subroutine.
Moreover, depending on which options you have activated, you should modify a specific subroutine not the one you have changed.

Chi-Tuan