em_les

[廊桥遗梦]

As you can see in 1.TXT, I modified namelist, max dom=3 according to your method and the run was successful. However, it didn't generate the d03 file. Is this normal?How can I generate a d03 file?

 

[kwerner]

Hi, Can you package together all of your rsl.* files into a single *.tar or zipped file and attach that? Will you also issue the following in your wrf running directory and then also attach the ls.txt file? Thanks!

ls -ls >& ls.txt

 

[廊桥遗梦]

Thanks,sir! This is the files I packed as you said.

 

[kwerner]

Hi,
I think the issue may be that you have the following start times for your 3 domains:

domain 01 : 0001-01-01_00:00:00
domain 02 : 0001-01-01_01:30:00
domain 03 : 0001-01-01_00:00:00

And since the simulation is only set for 1 hour, it's never getting to the start time for d02. d03 won't be able to run without the conditions for d02. Try setting all domains to the same start time and re-run the test. Let me know if that makes any difference. It's okay to start a nest at a later time, but it can't be before the start time of its parent.

 

[廊桥遗梦]

I made revisions according to your suggestions, yet the d03 file still failed to be generated.

 

[kwerner]

Hi,
The namelist.input file you attached to your last post shows that d02 still doesn't have the same start time as d01 and d02. And d03 has a end time that is identical to the start time, so it's running for no time.

 start_year                          = 0001, 0001, 0001,
 start_month                         = 01,   01,   01,
 start_day                           = 01,   01,   01,
 start_hour                          = 00,   01,   00,
 start_minute                        = 00,   30,   00,

 end_year                            = 0001, 0001, 0001,
 end_month                           = 01,   01,   01,
 end_day                             = 01,   01,   01,
 end_hour                            = 01,   02,   00,
 end_minute                          = 00,   30,   00,

You need to change these values to the following:

 start_year                          = 0001, 0001, 0001,
 start_month                         = 01,   01,   01,
 start_day                           = 01,   01,   01,
 start_hour                          = 00,   00,   00,
 start_minute                        = 00,   00,   00,

 end_year                            = 0001, 0001, 0001,
 end_month                           = 01,   01,   01,
 end_day                             = 01,   01,   01,
 end_hour                            = 01,   01,   01,
 end_minute                          = 00,   00,   00,

 

[廊桥遗梦]

Sorry,I uploaded the wrong namelist.input file before. Could you pleasde help me check what the problem is now?

 

[kwerner]

Hi, Thanks for updating that. After looking closer at the sizes of your domains, I realized that the sizes of d02 and d03 are too large to fit inside their parent domains. By default, there needs to be at least 5 parent grid spaces between it's boundary and the child's boundary. Even though the parent_grid_ratio is 3:1, if you do the math, a domain with size 100x100 will not fit in its parent (40x40). You'll have to adjust the domain sizes if you want this to run.

Even so, I'm not sure it's even reasonable to simulate domains with resolutions as high as domains 02 and 03 are using. Because the resolution of d01 is 0.1 km, with a parent_grid_ratio of 3:1, that means the resolution for d02 is 0.033 km and d03 is 0.011 km, which is much higher than anything we've seen anyone use with WRF. If you're able to get it to work, and the results look reasonable to you, then that's great, but you'll need to check closely. It may just be better to simulate a single high-resolution domain, which is what is recommended for this idealized case.