undefined_value in WRF_noahMP

[xiao.zh]

Hi all,
When running the WRF scheme, I found that variables related to LAI, including TGV, TAH, TZV, etc., had abnormal values. After investigation, I determined that these areas had a land use type of 15 in MODIS, which corresponds to snow and ice.

By checking the source code, I discovered that when the land use type is 15, the LAI is set to undefined_value, and in the program, undefined_value=-1.E36, which corresponds to the numbers starting with -99999 shown in the figure. I want to know why the program sets undefined_value to such a large number. Can undefined_value be set to a smaller value, such as -0.0001?

Additionally, there are many places where my LU_INDEX=15, but only some grid points have an undefined_value for LAI. What could be causing this?

Thanks and look forward to your reply.

Xiao

 

[haiqingsong]

Hi all,
When running the WRF scheme, I found that variables related to LAI, including TGV, TAH, TZV, etc., had abnormal values. After investigation, I determined that these areas had a land use type of 15 in MODIS, which corresponds to snow and ice.

By checking the source code, I discovered that when the land use type is 15, the LAI is set to undefined_value, and in the program, undefined_value=-1.E36, which corresponds to the numbers starting with -99999 shown in the figure. I want to know why the program sets undefined_value to such a large number. Can undefined_value be set to a smaller value, such as -0.0001?

Additionally, there are many places where my LU_INDEX=15, but only some grid points have an undefined_value for LAI. What could be causing this?

Thanks and look forward to your reply.

Xiao

Which WRF version did you use?

 

[Ming Chen]

If you set rdlai = .false., can you see the similar issue?

 

[xiao.zh]

Which WRF version did you use?

Thanks for your reply. My WRF version is 4.6.0.

 

[xiao.zh]

If you set rdlai = .false., can you see the similar issue?

Hi Ming, thanks for your reply. Do you mean rdlai2d = .false., ? The issue came out without setting rdlai2d, I saw that default value of rdlai2d if false. Then I tried rdlai2d = .false. and rdlai2d = .true., including different dveg values, the results still had the issue.

I also tried to modify the original code. I changed the undefined_value to 0. and -0.0001 in two cases. I checked my result just now, the value were 0. and -0.0001 where the value were -99999* before. I am not sure whether this modification have impact on the whole NoahMP scheme. Do you have any suggestions?

Xiao

 

[Ming Chen]

Sorry for the typo. Yes the option should be rdlai2d.

In your namelist.input, rdlai2d = .true., indicating that LAI is from the static input data. Note that WPS won't allow undefined values and I believe that LAI in your met_em and wrfinput should be fine.

Can you upload your namelist.wps for me to take a look?

 

[xiao.zh]

Sorry for the typo. Yes the option should be rdlai2d.

In your namelist.input, rdlai2d = .true., indicating that LAI is from the static input data. Note that WPS won't allow undefined values and I believe that LAI in your met_em and wrfinput should be fine.

Can you upload your namelist.wps for me to take a look?

Hi Ming, please see the attached file for my namelist.wps.

 

[Luiz Candido]

The resolution of your geogrid data (geog_data_res = ' ' ) is not defined, so wps cannot find the database lai it should read.

 

[William.Hatheway]

Hi Ming, please see the attached file for my namelist.wps.

looking in the error it looks like geog_data_res = ' '

Try putting in 'default'