Running WRF-Chem with resolution less than 1 km

[Alice Burton]

Dear all,

I wonder if this is possible to run the WRF-Chem model with resolution less than 1 km? If yes, would you please help me with this issue?

Many thanks in advance.

Kind regards,

[Mary Barth]

Hi Alice,

Of course it's possible to run WRF-Chem at dx<1 km. I'm not sure what kind of configuration you are trying to achieve (e.g. dx: 100-1000m and real meteorology, or dx 100 m or less and run as an LES), so it is a bit hard to give you all the help you might need.

Setting up WRF for these types of simulations should be documented in the WRF help pages.

For the chemistry, the main thing is that the emissions need to be interpolated to grids < 1 km. Be sure to check whether the emissions tool can do that or not. 

For real meteorology, it's best to run a few nests to go from regional scale to < 1km. That way the boundary conditions will match the grid size better.

Good luck,

Mary

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

Mary Barth 

Senior Scientist, NSF National Center for Atmospheric Research   

phone: 303-497-8186    email: bar...@ucar.edu

Staff web page  SIMA Web Site  MUSICA Web Site

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

[Alice Burton]

Dear Mary,

Thank you very much for your helpful response and for pointing me toward the relevant considerations.

I am planning to set up three nested domains with a target horizontal resolution of around 300 meters. The main objective of my study is to investigate the transport and dispersion of traffic-related air pollutants in a mountainous region.

I am aware that modeling over complex terrain requires particular care in selecting physical and chemical parameterizations, and that at such high spatial resolution the model stability becomes even more sensitive. In addition, I understand that special attention is needed to ensure both dynamical and chemical stability when running WRF-Chem at this scale.

Given these aspects, I would greatly appreciate your advice on an appropriate modeling strategy for such a configuration, particularly in terms of nesting approach, physics options, and any stability-related considerations you would recommend based on your experience.

Your expertise would be extremely valuable for this work, and I would be very grateful for any guidance you could share.

Kind regards,
Alice

[Mary Barth]

Hello Alice,

Sounds like a great, yet challenging simulation! It should produce very interesting results. 

Here's my advice:

- Work with WRF or WRF-Chem with chem_opt=0 first to get a good meteorology representation with your nests. That way you can work out instabilities or other issues. Hopefully there is guidance on the WRF pages about simulating mountain-valley regions and what may be the right model configuration. If not, then look for past studies in the literature and see if there are examples there.

- Once you have a good meteorology simulation, then turn on the chemistry

- As I mentioned, the emissions will need to be down-scaled to 300 m grids and it is good to run chemistry in all 3 domains so that the mesoscale motions can provide more realistic spatial distributions in the 300-m domain. 

- The other things to think about are the time steps for the chemistry processes. With a 300-m domain the time step for dynamics is likely 1-5 seconds. You probably don't need to calculate chemistry that often, but that depends a little on how rapidly the physics and chemistry is changing. 

One more thought. You can also run WRF-tracer, if you are not interested in the chemical evolution, as there is a way to initialize tracers and provide your own emissions of tracers. This might be a nice option in the first step of your work. See Pfister et al. (2017) for an example of using WRF-tracer where the tracers are emitted according to an anthropogenic emissions inventory. 

- Mary

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

Mary Barth 

Senior Scientist, NSF National Center for Atmospheric Research   

phone: 303-497-8186    email: bar...@ucar.edu

Staff web page  SIMA Web Site  MUSICA Web Site

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

[Carl Drews]

Alice -

If you don't mind sharing; where are these mountains? I am asking because ACOM has developed a 3D visualization of chemical plumes that works with Google Earth:
3D animation of wildfire plumes
https://opensky.ucar.edu/islandora/object/conference%3A3513

At the finer model resolutions, the isosurfaces will interact with terrain. As an experiment: we could animate your model results (when they are available), and see what the 3D isosurfaces look like.

Carl

--

Carl Drews

Software Engineer / Web Operations
NSF NCAR | Research IT | ACOM

Pronouns: he, him, his
https://staff.ucar.edu/users/drews

[Naser Mohammadzadeh]

Dear Alice,

Hi,

I ran the WRF model with three nested domains (300 km, 1.5 km, and 100 m, with parent grid ratios of 1:5:3). Although my main study is based on WRF-Chem, I first ran the model to check the meteorological conditions. For the planetary boundary layer (PBL), I used the YSU scheme for the parent domain and LES for the two inner domains.

Based on my experience, it’s important to first ensure that your meteorological setup is well configured. I recommend running the model with WRF only (as Mary suggested, set wrf-chem = 0) to validate the meteorology. After that, you can enable chemistry and proceed with WRF-Chem simulations.

Another point to consider is the emission resolution, which is typically quite coarse (e.g., 0.1° × 0.1°). For resolutions finer than 1 km, I’m not entirely sure how this will affect the results.

Could you please let me know which chemistry scheme you are using in WRF-Chem (i.e., the chem_opt option)?

Best regards,
Naser

On Tue, Apr 21, 2026 at 10:46 PM Alice Burton <alicebu...@gmail.com> wrote:

Dear Naser,

Thank you so much for your advice. But technically I think I should take some considerations to ensure the stability of the model both physically and dynamically. Have you ever done this before?

Any suggestion or advice is strongly appreciated.

Kind regards,

On Mon, Apr 20, 2026 at 1:55 PM Naser Mohammadzadeh <naser.moham...@gmail.com> wrote:

Dear Alice,

Technically, you can proceed in that way. However, first you need to generate the met_em files based on your chosen resolution.

After that, you should modify your namelist.input according to your WPS namelist.wps settings.

If you are using a single domain, the setup is relatively straightforward. However, if you are working with nested domains, you will need to carefully choose an appropriate grid ratio between the parent and nested domains.

Good luck with your setup.

Kind regards,

Naser

To unsubscribe from this group and stop receiving emails from it, send an email to wrf-chem+u...@ucar.edu.

[Alice Burton]

Dear Mary,

Thank you very much for your helpful and detailed guidance. I truly appreciate you taking the time to share these insights—they give me a much clearer direction on how to proceed with the simulation.

Your suggestions about starting with meteorology, carefully handling emissions scaling, and considering WRF-tracer are especially valuable. I will follow this approach as I move forward with my work.

I may have more questions as I progress, and I would really appreciate the opportunity to reach out to you again and benefit from your expertise.

Thank you once again for your support.

Best regards,
Alice

[Alice Burton]

Dear Carl,

Thank you very much for your message and for sharing this interesting resource. I really appreciate your suggestion—it sounds like a great idea to visualize the model results in 3D.

The study area is located over the Alps mountains.

I would be very interested in trying this approach once the model results are available, as it could provide valuable insights into how the isosurfaces interact with the terrain.

Thanks again for your support and suggestion.

Best regards,
Alice

[Alice Burton]

Dear Naser,

Thank you very much for your helpful guidance and for sharing your experience. I really appreciate your suggestions.

I am quite new to this field, so your advice is very valuable to me. I wanted to ask—could you please explain why you used the LES scheme for the inner domains?

Regarding the chemistry scheme (chem_opt), I have not made a final decision yet. However, I am inclined to use a scheme that includes secondary organic aerosols.

Thank you again for your support.

Best regards,
Alice