Fatal Error: VEGEFLUX

[Berina Mina Kilicarslan]

Dear All,

I am using WRF-Hydro 5.1.1 in stand alone mode compiled with the SPATIAL_SOIL=1.

I have created the soil_properties.nc file by using the R script provided in the WRF-Hydro user page. Based on the given information, I specified soilFillVal to 6 (Loam) which is the most common soil type in my domain. However, I faced the error below. 

t0OutputFlag:            1

  CRITICAL PROBLEM: HCAN <= ZPD
  i,j point=           1          61
  HCAN  =   0.00000000
  ZPD   =   1.78331975E-02
  SNOWH =   1.78331975E-02

 FATAL ERROR: CRITICAL PROBLEM IN MODULE_SF_NOAHMPLSM:VEGEFLUX  

Any help would be greatly appreciated.

Best regards,

Berina  

--

Berina Mina Kilicarslan

+90-312 2105415

Middle East Technical University

Civil Engineering Department

Water Resources Lab. (K4)

Ankara, Turkey

[Somaye Imani]

hi  Berina ,

I have the same problem. Can you solve it? 

[Berina Mina Kilicarslan]

Hi Somaye,

 

Unfortunately, couldn’t solve it yet.

 

 

Sent from Mail for Windows 10

[Katelyn FitzGerald]

Hi Berina and Somaye, 

The errors you're seeing are coming from Noah-MP and likely related to parameters associated with the land cover types (the model is seeing a canopy height of zero over a supposedly vegetated class).

Which land cover and soil classes are you working with?  

[Berina Mina Kilicarslan]

Hi Katelyn,

I am working with "modis_landuse_20class_30s". 

At that time, I noticed that the create_SoilProperties.R code was reading the USGS part of the MPTABLE.TBL. I modified the code so that it was reading the MODIS part, then I re-created the soil_properties.nc file. But, still facing the same problem. 

Thank you.

Katelyn FitzGerald <kate...@ucar.edu>, 28 Eyl 2020 Pzt, 19:04 tarihinde şunu yazdı:

[Somaye Imani]

Hi Katelyn,

Thanks for your reply. As berina said,  I am working with "modis_landuse_20class_30s" and using the create_SoilProperties.R script that reading the USGS part of the MPTABLE.TBL specialized for Version 5.1.

I am working with WRF data as forcing that used the Noah model for modeling the LSM.

--


You received this message because you are subscribed to a topic in the Google Groups "wrf-hydro_users" group.


To unsubscribe from this topic, visit https://groups.google.com/a/ucar.edu/d/topic/wrf-hydro_users/GhdzCPLIVzI/unsubscribe.


To unsubscribe from this group and all its topics, send an email to wrf-hydro_use...@ucar.edu.


To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/2f6ea632-16d2-4a70-a6f7-c610d95a7428n%40ucar.edu.

--

Sincerely,
Somaye Imani
Research fellow at Water Research Institute (WRI)

And

Ph.D. Candidate, Environmental Engineering, University of Tehran, Iran

[Somaye Imani]

Dear  Katelyn, 

I still can not solve this problem.  I'm working on several projects with WRF-Hydro and face this problem randomly in some cases.  In the latter case, the LU_Index of USGS was modified in the geo_grid.nc file. In some modified land-use classes like water bodies, parameters such as HVB (Bottom of canopy) and HVT (Top of canopy) are zero. I changed these parameters but it is not fixed the problem and the model message Error in MODULE_SF_NOAHMPLSM: VEGEFLUX with zero in HCAN, ZDP, and SNOWH again. I could not find the method of calculating the HCAN, ZDP, and SNOWH parameters in the model.

How to calculate these parameters in the model? Based on which parameter and input files to the model are these parameters calculated?

I would appreciate it if you could give me more details.

Best regards,  
Somaye

[Katelyn FitzGerald]

Unfortunately the create soil properties script isn't currently set up to work with the MODIS land classes (only USGS).  There's an update in progress, but I'm not sure what the timeline is for that.

Particularly if you're editing scripts or files yourself, you'll want to make sure you're using consistent land use / cover classes and land / water masks across your files (e.g. soil properties, geogrid, wrfinput, etc.).

[Somaye Imani]

Dear Katelyn,

Thank you for the information. I created soil_properties.nc with USGS land use data. Also, all input and modified files are in the same Domain. unfortunately, I face this error again.

Sincerely,
Somaye Imani
Research fellow at Water Research Institute (WRI)

And

Ph.D. Candidate, Environmental Engineering, University of Tehran, Iran

[Yiannis]

Dear all,

Have you found a way to solve the issue?

I have tried running WRF-H v5.1 and I get the following, which stops my simulation:

 CRITICAL PROBLEM: HCAN <= ZPD

  i,j point=          78          13

  HCAN  =  0.0000000E+00

  ZPD   =  0.0000000E+00

  SNOWH =  0.0000000E+00

wrf_abort

  CRITICAL PROBLEM: HCAN <= ZPD

  i,j point=         120          48

  HCAN  =  0.0000000E+00

  ZPD   =  0.0000000E+00

  SNOWH =  0.0000000E+00

wrf_abort

Did you need to change the land use dataset you use?

In my case, I had originally the MODIS land use dataset as I got it from my geogrid file from WPS v4. I replaced this dataset with another land use dataset (Corine Land Cover) that I reclassified to MODIS classes.

Are the MODIS land use classes really the problem?

Do I need to reclassify my Corine Land Cover to the USGS classes instead of to MODIS?

Thank you,

Yiannis

[Katelyn FitzGerald]

The MODIS land classes shouldn't be an issue for the model, but there can be some problems w/ inconsistent land cover masking and/or missing attributes (e.g. for water cells, etc.) and the create_soilproperties.R script doesn't currently support the MODIS categories so these are common points of failure. 

[Yiannis]

Dear Katelyn,

Thank you very much.

I tried to solve my problem in two ways:

(1) I modified two lines in the  create_soilproperties.R in order to correctly read the parameters for the MODIS classes instead for the USGS classes. The script completed successfully, so I guess the problem isn't with the script itself that doesn't support the MODIS categories.  But still I got the same error after running the wrf hydro executable. Similar to what Berina reported on the 28th of September in this discussion. 

(2) I went back to WPS and re-created my geo_em.d03.nc file with USGS land use data this time, instead of the default MODIS data for WRF V4. I ran the create_soilproperties.R script and then the WRF-Hydro executable. Unfortunately I got the same errors even for the USGS land use classes, similar to what Somaye reported on the 7th of November in this discussion.

Please find attached the layer of hvt, which is the HCAN variable in NOAH-MP from where we get the error :  CRITICAL PROBLEM: HCAN <= ZPD
You can see that the layers for hvt variable from both the MODIS and the USGS datasets look fine. I wouldn't expect that our datasets have issues that cause the error. 

Could there be something that doesn't translate correctly from the soil_properties.nc in the Noah-MP part of the code of WRF-Hydro?

  

Thank you,

Yiannis

[Katelyn FitzGerald]

Yes, it's likely an issue related to the land masking in the model resulting in the HCAN value of 0.  Unfortunately there are several places this information is pulled from in the model so it can be a bit difficult to diagnose.  We do have working test cases using both USGS and MODIS land cover classes so it's unlikely, but possible that you've hit some sort of edge case that didn't arise there.

One more thing to check quickly... have you by chance also updated the wrfinput file(s) to be consistent with your new geogrid and soil_properties files?

If you're still having trouble feel free to follow up with an example case to the WRF-Hydro contact form and we can try to help diagnose further there and follow up w/ relevant notes here for other folks.

[Yiannis]

Hi,

I would like to let you know that I have solved the issue that was causing the error  CRITICAL PROBLEM: HCAN <= ZPD .

Your tip Katelyn for checking the wrf_input file was the one that did help me (so thanks for that). To initialize the WF-Hydro, I was using the wrf-input file as I got it from the pre-processing for my WRF simulations. 

But then I also tried to generate a new wrf_input file using the create_wrfinput.R script. This new wrf_input file caused no error in the simulation. 

/Yiannis

[Katelyn FitzGerald]

Glad to hear and thanks so much for following up!  

[Samar Minallah]

Hi Katelyn,

I have run into the same error and have not been able to resolve based on the solutions in this discussion:  

CRITICAL PROBLEM: HCAN <= ZPD

  i,j point=         356         429

  HCAN  =  0.0000000E+00

  ZPD   =  0.0000000E+00

  SNOWH =  0.0000000E+00

I am using NAM forcing data, the geogrid is 3km with USGS land cover/soil classes from the default table, and routing grid 300m. The WRFHydro version is 5.1.1. When I do the same runs at coarse resolution (geogrid 8-9 km), everything works fine, but as a I move to finer resolution this error pops up. I have also used both create_wrfinput.R and WRF real to create the initial LS conditions file but run into the same issue at finer resolutions.

Do you have any other insight into this issue and how to troubleshoot it.

Thank you,

Samar

[Katelyn FitzGerald]

Hi Samar,

It's likely there's some sort of inconsistency in the land cover and/or land masks in your finer resolution geogrid, soil properties (if this is something you're using), and/or wrfinput files.  

If you're using a recent version of WRF to generate your geogrid and/or wrfinput files, the default land cover dataset has changed to MODIS.  Is it possible you've accidentally used the MODIS classes?  You should be able to check by dumping the metadata in the file header (i.e. "ncdump -h FILENAME" on a machine w/ the netcdf libraries installed).

If that's not the case, I would start by checking the land masks and classifications in the files mentioned above for consistency and/or unrealistic inputs.  

If you're still having trouble, feel free to follow up with an example case to the WRF-Hydro contact form and we can try to help diagnose further.  

--

Katelyn FitzGerald (she/her/hers)

NCAR Research Applications Laboratory

kate...@ucar.edu

[ershier]

hi  Yiannis ,

I have the same problem.   I checked the wrf_input file but it's fine.Could you tell me how you solved the problem? 

[aubrey]

The things you should check are:

1) Run with idealized forcings. If you do not get this error then the issue is with your forcing files (likely missing data).

2) Run with SPATIAL_SOIL turned off. If you do not get the error then the issue is with your 2d/3d soil parameter file.

3) Make sure you are using the latest create_SoilProperties script. We have versions in both R and python on our website that you can try. Check your value ranges in the file you generate and make sure they look reasonable (e.g., no missing values). Make sure your land/water distinctions are consistent across all your files.

Hope that helps.

Aubrey