Why there are differences in landuse between wrfinput and wrfout file?

[xxi]

Hi,
I'm running em_les and trying to modify the ideal landuse by modifying codes. I have successfully achived (picture 1：LU_INDEX in wrfinput file, the colors represent different landuse type, 1 and 11 respectively) ,but when I run wrf.exe, I finally get picture 2 (LU_INDEX in wrfout file, here even the value 0 exists, and the distribution is different). There is not any error in the whole process and I'm really confused.
I'd appreciate it if you have any idea.

 

[kwerner]

Hi,
Are you turning on any physics options when you run wrf? Can you attach your namelist.input file?

 

[xxi]

Hi,
Are you turning on any physics options when you run wrf? Can you attach your namelist.input file?

Hi,
I turned on land surface and surface layer physics in wrf-les. Besides, when I did the same with different land distribution, it worked well.

 

[kwerner]

Besides, when I did the same with different land distribution, it worked well.

Does this mean that you are able to move past this issue with a different land distribution, or do you still need help with this one?

 

[xxi]

Does this mean that you are able to move past this issue with a different land distribution, or do you still need help with this one?

I still need help with this one because I need this specific land distribution. I just put in the extra finding before so that it might be useful.

 

[kwerner]

Apologies for the delay. Can you use the attach option to attach your full namelist.input file, as well as the modified code file(s)? Please also let me know which version of WRF you're using. Thanks!

 

[xxi]

Apologies for the delay. Can you use the attach option to attach your full namelist.input file, as well as the modified code file(s)? Please also let me know which version of WRF you're using. Thanks!

Hi,
I'm sorry I didn't noticed your reply before.I'm using WRF4.3,and all of my modification is attached: revision1 is in dyn_em/module_initialized_ideal.F, revison2 is in phys/module_surface_driver.F.
Thanks very much!

 

[kwerner]

Hi,
Thanks for sharing those. I'm not certain I made the exact right modifications based on the revision files you sent. I had to kind of guess where those revisions should go in the code. However, I recompiled the code and ran the case with your namelist. My landuse plots in wrfinput_d01 and wrfout_d01* don't look like yours, and they are identical. See the attached pic.

Just as a test, can you try this with WRFV4.7.1 to see if that makes any difference? There have been quite a few modifications to the surface driver since V4.3. I'm not sure if those changes would affect your simulation, but it would be nice to know.

If that makes no difference, can you let me know if you get the same issue even if you don't modify the code at all - just using "out-of-the-box" code? Thanks!

 

[xxi]

Hi,
Thanks for sharing those. I'm not certain I made the exact right modifications based on the revision files you sent. I had to kind of guess where those revisions should go in the code. However, I recompiled the code and ran the case with your namelist. My landuse plots in wrfinput_d01 and wrfout_d01* don't look like yours, and they are identical. See the attached pic.

Just as a test, can you try this with WRFV4.7.1 to see if that makes any difference? There have been quite a few modifications to the surface driver since V4.3. I'm not sure if those changes would affect your simulation, but it would be nice to know.

If that makes no difference, can you let me know if you get the same issue even if you don't modify the code at all - just using "out-of-the-box" code? Thanks!

Hi,
Thanks for your patient reply! Actually, I sometimes success like yours but sometimes fail, even though I just repeat without any modification.It's a kind of probability for me. (Ignore the specific distribution pattern, I may have given the code with uniform distribution instead of strip distribution by mistake. I just want to emphasize the problem that the distribution makes shift towards the right in my original picture.) I heared that it may related to the number of cores or some process in parallel computing and I'm not sure.
As for the version you mentioned, I think it maybe the reason and 'll try later. Thank you again!