MOZBC with nested simulation

[Jaya Thakur]

Dear Gabriele/Stacy/Rajesh,

I faced a problem when trying to run for a nested domain. I have one parent (d01) and two nested domains(d02 and d03) in my simulation. I have 4 files generated (wrfbdy_d01, wrfinput_d01, wrfinput_d02, wrfinput_d03) on execution of real.exe with the chemistry on.

If my mozbc input file has the following settings:

Case 1:

do_ic = .true.

do_bc = .true.

domain = 3

Result: Only the wrfinput_d03 is updated (I have checked this with the time of file written, because the size of the file did not change)

Case 2:

do_ic = .true.

do_bc = .true.

domain = 2

Result: Only the wrfinput_d02 is updated (I have checked this with the time of file written, because the size of the file did not change)

Case 3:

do_ic = .true.

do_bc = .true.

domain = 1

Result: Both the wrfinput_d01 and wrfbdy_d01 are updated (I have checked this with the time of file written, because the size of the file did not change)

Clarifications needed:

1. From the README file of MOZBC, I understood that the variable "domain" is to specify the total number of WRF domains, but with the above trials, it looks as if it is actually the domain id. Kindly clarify.

2. It seems from my cases, that all the domains boundary and initial conditions cannot be updated with only one simulation. Should the mozbc be run for all the domains separately..

3. You have suggested to make (in your communication the do_bc=.false for Domain 2, but since the wrfbdy_d02 is anyway not available in the WRF directory, it wouldn't update that. Is there any issue running it like mentioned in case 2.

Thanking you in advance.

Best regards

Jaya Thakur

[Gabriele Pfister]

Hello Jaya, 

see below my response to your questions.

Gabi

On Thu, Jul 29, 2021 at 2:30 AM Jaya Thakur <jaya...@gmail.com> wrote:

Dear Gabriele/Stacy/Rajesh,

I faced a problem when trying to run for a nested domain. I have one parent (d01) and two nested domains(d02 and d03) in my simulation. I have 4 files generated (wrfbdy_d01, wrfinput_d01, wrfinput_d02, wrfinput_d03) on execution of real.exe with the chemistry on.

If my mozbc input file has the following settings:

Case 1:

do_ic = .true.

do_bc = .true.

domain = 3

Result: Only the wrfinput_d03 is updated (I have checked this with the time of file written, because the size of the file did not change)

Case 2:

do_ic = .true.

do_bc = .true.

domain = 2

Result: Only the wrfinput_d02 is updated (I have checked this with the time of file written, because the size of the file did not change)

Case 3:

do_ic = .true.

do_bc = .true.

domain = 1

Result: Both the wrfinput_d01 and wrfbdy_d01 are updated (I have checked this with the time of file written, because the size of the file did not change)

Clarifications needed:

1. From the README file of MOZBC, I understood that the variable "domain" is to specify the total number of WRF domains, but with the above trials, it looks as if it is actually the domain id. Kindly clarify.

This is correct. You set the domain for each domain separately. 

2. It seems from my cases, that all the domains boundary and initial conditions cannot be updated with only one simulation. Should the mozbc be run for all the domains separately..

See 1.

3. You have suggested to make (in your communication the do_bc=.false for Domain 2, but since the wrfbdy_d02 is anyway not available in the WRF directory, it wouldn't update that. Is there any issue running it like mentioned in case 

I have not tried it out but believe mozbc would fail because it does not find the file. So for the inner domains, simply set do_bc=.false