How Does WRF-CMAQ Handle NOx Emissions Over Time Steps?

I’ve observed that when I double the NOx emissions for a specific hour, the resulting change in NO2 columns for the next hour is quite small (the NMB to pseudo-observations is less than 2% across the US). I’m particularly interested in understanding the model’s treatment of emissions over time steps. Are hourly NOx emissions released anew at the beginning of each hour? Furthermore, to what extent do emissions from previous hours continue to influence the simulation?


Could you provide more information about the model versions you are using? Hourly emission inputs (e.g., NOX, SO2, etc.) are interpolated to the model step and vertically mixed in the vertical diffusion model, before processed in other science modules (advection, chemistry, aero, etc).

When looking at emission impacts on NO2 columns, please note, that the NO2 concentration within the column can be quite large, so seeing emission perturbations within that column can be hard. I suggest looking at the layer you did the emission scaling first. For example, if you scaled surface (layer 1) NOX emissions by 2, my first order analysis would be to look at the surface (layer 1) NO2. Lastly, looking at the NO2 concentrations in the CONC/ACONC files may not be a very good proxy. Before those values are written, remember, the model goes through several source/sink processes that will impact NO2 concentrations.

Emissions from previous hours, may continue to influence the simulation, but depends on many factors (domain size, strength of source, etc.) so it’s hard to say.

Hope this help!

1 Like

I absolutely agree all these comments. Just an additional note that it is always good to activate the PA tool to capture and track the changes of NOx emissions (also an effective method for confirming the emission change has been implemented).

1 Like

Hi Fahim,

Thank you for your help! I’m using CMAQv5.4, with emissions sourced from EQUATES and covering a grid resolution of 12 km within the 12US1 domain. I did the same emission scaling factor for all vertical layers, and analyzed the impact on the surface layer (~15%) and the whole columns (~2%). I understand that the model goes through source/sink processes before written to CONC/ACONC files. Could you provide more options If solely focusing on NO2 concentrations in CONC/ACONC files is not recommended?

I appreciate your insights.

Thank you for the suggestion! I’ll try it.


Expanding on the comments made by @Ryan I suggest you print out DESID Diagnostic files during your run. The instructions to do so are found here: CMAQ/DOCS/Users_Guide/Appendix/ at main · USEPA/CMAQ · GitHub.

This will quickly give you a way to visualize if your scaling instructions were interpreted correctly by DESID. If you scaled things correctly, in the DESID diagnostic file you should see the scaled amount of NOx is the scaled amount above the input emissions for that stream.

1 Like