Re: [wrf-hydro_users] Digest for wrf-hydro_users@ucar.edu - 2 updates in 1 topic

[Yongxin Zhang]

HI Fatemeh,

     Yes, we can start from scratch and create the NLDAS-based forcing for your domain. I am not sure if you want to start from the examples that we have provided as part of the model code release or if you want to start from the real NLDAS data that will need to be downloaded.

      You can find the regridding package for NLDAS (ESMFregrid_NLDAS.tar_.gz) under:

       https://ral.ucar.edu/projects/wrf_hydro/pre-processing-tools

and once you unpack this package, you will find the following files and sub-directory together with the READM.txt file:

1) geo_em.d01.nc: the sample WRF-Hydro domain file;

2) NLDAS2WRFHydro_generate_weights.ncl: the NCL script that creates the weight files for mapping between the NLDAS grid and the WRF-Hydro grid. This needs to run once since the weight file won't change as long as the WRF-Hydro domain does not change;

3) NLDAS2WRFHydro_regrid.ncl: the NCL script that does the regridding using the weight file that was created above. Optionally, we can also use this script to downscale temperature, humidity, and surface pressure to account for terrain difference between the NLDAS grid and the WRF-Hydro grid but there is some pre-processing involved in doing this.

4) input_files: the sample NLDAS files that needed to be regridded onto the WRF-Hydro domain.

       If you want to use the real NLDAS data, depending on your domain of interest, for the U.S. CONUS domain, you can first download the NLDAS data files from the following website:

      https://hydro1.gesdisc.eosdis.nasa.gov/data/NLDAS/NLDAS_FORA0125_H.2.0/

And then you can use the NLDAS2WRFHydro_generate_weights.ncl file to create the weight file and then use the created weight file to generate the WRF-Hydro forcing files using NLDAS2WRFHydro_regrid.ncl.

        Please let us know how you want to proceed with this.

Thanks,

Yongxin

On Fri, May 16, 2025 at 5:13 AM <wrf-hyd...@ucar.edu> wrote:

wrf-hyd...@ucar.edu

Google Groups

Topic digest
View all topics

• Digest for wrf-hyd...@ucar.edu - 1 update in 1 topic - 2 Updates

Digest for wrf-hyd...@ucar.edu - 1 update in 1 topic

Yongxin Zhang <zha...@ucar.edu>: May 15 10:28AM -0600 HI Fatemeh,   I can help you with generating the HRLDAS forcing files using NLDAS inputs but I am not not sure what you meant by the main configuration file. Are you referring to the two namelist files, namelist.hrldas and hydro.namelist, or the NCL script that is used to regrid and downscale the NLDAS data and create the HRLDAS forcing files?   Thanks, Yongxin

Fatemeh Shirkhanloo <fate...@pdx.edu>: May 15 09:50AM -0700 Dear Yongxin,   Thank you so much for your response.   I would like to start from scratch and follow your approach. Could you please guide me through the correct procedure for generating the HRLDAS forcing files using NLDAS inputs?   I really appreciate your help.   Best regards, Fatemeh Shirkhanloo

Back to top

You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to wrf-hydro_use...@ucar.edu.

[Fatemeh Shirkhanloo]

Dear Yongxin,

Thank you so much for your help.

I was trying to download the GRIB files as you mentioned, but unfortunately, it seems that the GSFC NASA website no longer provides GRIB files after 2024—they have converted them to NetCDF format.

Since the approach you shared appears to require GRIB files for use with the NCL scripts, I’m not sure how to proceed. What should I do in this case? Is there a way to work with the NetCDF files, or should I look for older GRIB data elsewhere?

I’d really appreciate your guidance on the next steps.

Best regards,
Fatemeh Shirkhanloo

[Yongxin Zhang]

Dear Fatemeh,

       Many thanks for letting me know about the changes in the NLDAS file format from GRIB to netCDF around 2024. I can help with the netCDF files as well. I will modify the file that creates the weight file and the file that performs the regridding so they can work with the netCDF files. I will also write a simple csh or bash script to call those NCL files so you just need to run the shell script. I will try to send all these to you within the next few days.

Thanks,

Yongxin

[Fatemeh Shirkhanloo]

Dear Yongxin,

I really appreciate your help and support.

Best regards, 
Fatemeh

[Yongxin Zhang]

Dear Fatemeh,

     You are very welcome.

Dear Caroline, 

      Thanks for your previous message on forcings. I will send you the NCL scripts and shell script for creating the NLDAS-based forcings as well. Due to the different types of forcing input files that we have worked with, it is not easy to create documentation or guidance to work with each type of forcing input files. However, we have developed a Python-based Meteorological Forcing Engine (MFE) specifically for the National Water Model (NWM) operational purposes and this MFE has been running operationally for a few years already. This MFE uses a config file to define the types of forcing input file and the options of regridding, downscaling, bias corrections, supplemental forcings, and other capabilities that are performed in Python modules/scripts. This MFE is easy to run and we are hoping to have this MFE released to the public by late this year or early next year.

Thanks,

Yongxin

On Wed, May 21, 2025 at 3:06 PM Fatemeh Shirkhanloo <fate...@pdx.edu> wrote:

Dear Yongxin,

I really appreciate your help and support.

Best regards, 
Fatemeh

[Yongxin Zhang]

Fatemeh, Caroline,

      I downloaded the NLDAS netCDF files for May 19, 2025 and modified the NCL scripts for working with these files and also created a simple csh script for creating the weight file and then doing the regridding. The csh script sets up the NCL environment (i.e., module load NCL) and then executes the NCL scripts. Please see the attached tar file named NLDAS_regridding.tar that also includes the WRF-Hydro domain file (geo_em.d01.nc) and the weight file. I also attached a tar file named input_files.tar that contains the input NLDAS netCDF files and another tar file named output_files.tar that contains the created NLDAS forcings for the WRF-Hydro domain. This regridding is a direct mapping from the NLDAS grid to the WRF-Hydro grid without accounting for the terrain difference between the two for temperature (and subsequently humidity and surface pressure) and terrain effect for radiation.

      Please let me know if you run into any issues

or have any questions.

Thanks,

Yongxin

 input_files.tar

On Tue, May 27, 2025 at 12:58 PM Caroline Hashimoto <cm...@rice.edu> wrote:

Hi Yongxin,

Thank you so much for the help! I'll try using the NCL regridding scripts.

Best,

Caroline

[Yongxin Zhang]

HI Fatemeh, Caroline,

    If you installed NCL using Conda, you will need to replace "module load ncl" with something like:

#source $MY_SW/miniconda3-amd64/bin/activate

#conda activate ncl_stable

Thanks,

Yongxin

[Fatemeh Shirkhanloo]

Dear Yongxin, 

I really appreciate your help. 
Finally I recreated HRLDAS forcing data for the Croton test case based on your scripts.

Best Regards, 
Fatemeh

[Yongxin Zhang]

Dear Fatemeh,

     It is good to know that you were able to create the HRLDAS forcing data for the Croton test case. Thanks for letting us know.

Yongxin

[Carrie Hashimoto]

Hi Yongxin,

Thank you for the help! I was able to regrid NLDAS files from here using the ESMF software after renaming the variables to match what is expected in the script and removing the time dimension, which the script doesn't expect the files to have.

Best,

Carrie

--

Carrie Hashimoto

Rice University 2025 | BSCE Civil Engineering, BS Statistics

(she/her) | 626-240-8535 | cm...@rice.edu

[Yongxin Zhang]

Hi Carrie,

      Thanks for letting us know!

Yongxin

[Yongxin Zhang]

Dear Fatemeh,

      You are very welcome!

      I am glad that you will be using CONUS404 as the forcing data. We are part of the broad CONUS404 team here at NCAR and we have been using CONUS404 to drive WRF-Hydro. I think that CONUS404 is a pretty good dataset.

      We have developed the Meteorological Forcing Engine (MFE) that directly ingests and processes CONUS404 for any domain over the CONUS but this MFE is still not in the public domain. We are currently working on the documentation and the public release of this MFE is scheduled for late this year or early next year so you won't be able to use it until it is publicly released.

      I can access CONUS404 on NCAR Research Data Archive (RDA) and I can download the wrf2d files that contain all the needed variables (I am wondering if you could double check to see if you could download the wrf2d files). I did not know that when you download CONUS404 from NCAR RDA each variable is written out in a separate NetCDF file. We can either first process each variable and then combine all the regridded variables together in the final LDASIN files or first combine all the variables together into one netCDF file and then process this single netCDF file but either way needs some coding work.

      If your test period is short, say a few days or a few weeks, I can create the LDASIN forcing data for the Croton domain using the MFE and then upload the files to a Google Drive folder that you can access. But if you are going to run the model for a longer time period, then we will need to find another way to deal with this.

      Please let me know.

Thanks,

Yongxin

On Thu, Jun 26, 2025 at 1:46 PM Fatemeh Shirkhanloo <fate...@pdx.edu> wrote:

Dear Yongxin,

I hope this email finds you well.

I truly appreciate the guidance and files you kindly shared with me previously, which helped me successfully convert NLDAS netCDF data into the HRLDAS forcing format. Your assistance was invaluable.

I am now working on running the Croton test case model that I built from scratch using CONUS404 forcing data. The CONUS404 data I downloaded from NCAR is already subset for the Croton domain. However, each variable is stored in a separate NetCDF file—for instance, wind speed and longwave radiation are provided in different files.

I would like to convert this set of files into the HRLDAS forcing format (Type 1) to enable running the model. I would greatly appreciate it if you could advise me on whether there is an approach or tool you recommend for handling this type of conversion.

I have attached a few sample NetCDF files for your reference.

Thank you very much for your time and support.

Best regards,
Fatemeh Shirkhanloo

[Yongxin Zhang]

Dear Fatemeh,

      I created the CONUS404-based LDASIN forcing files for the Croton domain for August 26, 2011 to September 2, 2011. Please see the attached CONUS404_forcing.tar file. The forcing files are hourly files and the files are named as YYYYMMDDHH00.LDASIN_DOMAIN1, i.e., with the extra "00" for representing minutes. This extra-minute designation is useful when you have sub-hourly input forcings such as 15-minute WRF output or 5-minute radar data. Because of these file names, you will need to use FORC_TYP = 2 instead of FORC_TYP = 1 in namelist.hrldas (there are no differences between 2 and 1 except that the LDASIN file names have minutes in them for 2). 

      Please let us know if you run into any issues with these files or have any questions.

Thanks,

Yongxin

On Fri, Jun 27, 2025 at 12:05 PM Fatemeh Shirkhanloo <fate...@pdx.edu> wrote:

Dear Yongxin,

Thank you so much for your detailed response and for your continued support. I truly appreciate the time and effort you dedicate to helping me. Your guidance has been invaluable throughout this process.

Regarding your kind offer, I would be extremely grateful if you could help me generate the LDASIN forcing data for the Croton test case using the MFE.

I plan to run the Croton test case using CONUS404 forcing data for a short period of one week, from August 26, 2011 to September 2, 2011, with hourly resolution. This run is primarily intended as a test to compare the model performance when driven by CONUS404 versus NLDAS forcing. Once I verify that everything works properly and assess the differences, I will shift my focus to my actual research domain, which is the Willamette River Basin. For that, I will generate a new domain and use CONUS404 to force the model over that region and obviously for a longer period of time.

Please let me know if it would be possible for you to help me with the LDASIN files for this short test period. I would greatly appreciate it.

Thank you once again for your time, generosity, and continuous support.

Best regards,
Fatemeh