ndown not working

[ukmetmodeller]

Hi,

I've previously run WRF with a parent (d01) and then 2 nested domains (d02 and d03).

However, I wish to rerun the 2nd nested domain (d03), the extents were not sufficient.

1) I've run WPS for domain 3 with the updated domain extents.
2) I've used the met_em files from the updated domain d03 and renamed them to domain d01 (via a system link) i.e. met_em.d03 to met_em.d01
3) I've updated the namelist.input file shifting domain parameters d03 to column 1.
4) I've run real.exe - successfully completed.
5) Renamed wrfinput_d01 to wrfndi_d02
6) linked the wrfout files from d02 (wrfout_d02*) and renamed (via system link) to wrfout_d01*
7) I've updated the namelist.input file shifting domain d02 to column 1 and d03 to column 2.
8) I've run ndown.exe.

This is where the issue starts. ndown.exe runs and doesn't flag any errors. It states that wrf task is running see below. The rsl.out.0000 and rsl.error.000 files show limited information (see below).The ndown.exe was left running for 72 hours to calculate 4 hours. For reference when running WRF with all three domains it took about an hour to run 4 hours. I've set the debug level in namelist.input to 100 and no additional information.

Is is possible to get some advice on why ndown might be stuck in an infinte loop with no messages?

>> mpirun -np 32 ./ndown.exe
starting wrf task 0 of 32
starting wrf task 1 of 32
starting wrf task 4 of 32
starting wrf task 5 of 32
starting wrf task 6 of 32
starting wrf task 8 of 32
starting wrf task 9 of 32
starting wrf task ...

FILE: rsl.out.0000
taskid: 0 hostname:
Ntasks in X 4 , ntasks in Y 8
*************************************
Configuring physics suite 'conus'

mp_physics: 8 8
cu_physics: 0* 0*
ra_lw_physics: 4 4
ra_sw_physics: 4 4
bl_pbl_physics: 2 2
sf_sfclay_physics: 2 2
sf_surface_physics: 2 2

(* = option overrides suite setting)
*************************************
Domain # 1: dx = 3000.000 m

FILE: rsl.error.0000
taskid: 0 hostname:
module_io_quilt_old.F 2931 T
Ntasks in X 4 , ntasks in Y 8
*************************************
Configuring physics suite 'conus'

mp_physics: 8 8
cu_physics: 0* 0*
ra_lw_physics: 4 4
ra_sw_physics: 4 4
bl_pbl_physics: 2 2
sf_sfclay_physics: 2 2
sf_surface_physics: 2 2

(* = option overrides suite setting)
*************************************
Domain # 1: dx = 3000.000 m

 

[kwerner]

Hi,
Can you tell me what you've changed about d03 in the WPS process? Are you using a different domain size, resolution, input data, etc.?

 

[ukmetmodeller]

Hi, thank you for replying.

I've sorted the issue now.

However, for completeness, in response to your question: the i j grid change so that the nest domain shifted.

 

[skarozis]

Hi,
Can you provide info how you sorted out the issue?
I have the same issue in namelists and domains I ve been usinf since WRFV3.4. Up to WRFV4.1.2 I had no issue but now with WRFV4.4.2 I am stuck in ndown with no error.

 

[ukmetmodeller]

Hi,
Can you provide info how you sorted out the issue?
I have the same issue in namelists and domains I ve been usinf since WRFV3.4. Up to WRFV4.1.2 I had no issue but now with WRFV4.4.2 I am stuck in ndown with no error.

Hi, sorry I missed this question and have only just spotted it.
So the problem I had was that the instructions I followed suggested changing the grid (column) from the child grid to the parent grid and a series of steps but the guide I followed used the child ID even when in the first column. In the end I just had to change the ID so the first column was 1 and then the next 2 etc like you would normally have for namelists. I'm not sure if this is a change in more recent versions or not but several guides showed the namelist with a non 1 ID in the first column so maybe it is. Hoepfully that helps.

 

[Ming Chen]

Please take a look at the document here:

Chapter 5: WRF Model

It describes in detail how to run ndown step by step.

 

[ukmetmodeller]

Please take a look at the document here:

Chapter 5: WRF Model

It describes in detail how to run ndown step by step.

Yes all good. Been running fine. The out of date tutorials I found online didn't have the parent ID changed to 1 when a 2nd or 3rd column was shifted to the first column. The tutorial doesn't mention the need for it to be able to run. As soon as I sorted that it was all fine.