Simulated plume re-renters the domain after some time??

[HP1801]

I have been trying to simulated a plume with the WRF_LES model, and I currently have 2 domains, one nested into another. I used a resolution of 10 m for my nested domain and 30 m for the larger domain. I am not sure why the plume re-enters the domain after a little while (image below)

 

[Ming Chen]

Would you please give me more information about this case, like the namelist.input and how you initialize the plume. The plot is for domain 2, I guess?

 

[HP1801]

Would you please give me more information about this case, like the namelist.input and how you initialize the plume. The plot is for domain 2, I guess?

Yes the plot is for domain 2. I have attached the namelist.input, module_initialize_les.F and solve_em.F for further information.

More specifically, in module_initialize_les.F, I added the following lines to initialize the plume:

   IF (config_flags%tracer_opt .eq. 2) THEN
     DO j = (jde + jds)/2, (jde + jds)/2, 1
       DO i = 3*(ide + ids)/4, 3*(ide + ids)/4, 1
          IF ( its .LE. i .and. ite .GE. i .and. jts .LE. j .and. jte .GE. j ) THEN
            tracer(i, 1, j, P_tr17_1) = 0.
!                tracer(i, 1, j, P_tr17_2) = 1.
!                tracer(i, 1, j, P_tr17_3) = 1.
!                tracer(i, 1, j, P_tr17_4) = 1.
!                tracer(i, 1, j, P_tr17_5) = 1.
!                tracer(i, 1, j, P_tr17_6) = 1.
!                tracer(i, 1, j, P_tr17_7) = 1.
!                tracer(i, 1, j, P_tr17_8) = 1.
          END IF
       END DO
     END DO
   END IF

And in solve_em.F, I added the following lines:

   IF (config_flags%tracer_opt .eq. 2) THEN
     DO j = (jde + jds)/2, (jde + jds)/2,  1
       DO i = 3*(ide + ids)/4, 3*(ide + ids)/4, 1
          IF ( ips .LE. i .and. ipe .GE. i .and. jps .LE. j .and. jpe .GE. j ) THEN
            tracer(i, 1, j, P_tr17_1) = 1.
!                tracer(i, 1, j, P_tr17_2) = 1.
!                tracer(i, 1, j, P_tr17_3) = 1.
!                tracer(i, 1, j, P_tr17_4) = 1.
!                tracer(i, 1, j, P_tr17_5) = 1.
!                tracer(i, 1, j, P_tr17_6) = 1.
!                tracer(i, 1, j, P_tr17_7) = 1.
!                tracer(i, 1, j, P_tr17_8) = 1.
          END IF
       END DO
     END DO
   END IF

 

[Ming Chen]

The modified code in solve_em.F indicates that at each time step, the tracer will be released at the specified points in both parent and child domain. In this case, it is hard to tell whether the tracer is newly released or it is reentering the domain.