Big Contribution from BCON to NO3 component of PM2.5


I have completed a series of sensitivity analyses using CMAQ-ISAM version 5.4, focusing on various emissions scenarios. The simulations were run across three nested domains: the third domain encompassing the entirety of the UK and adjacent shipping areas, and the parent domains extending over most of Europe and North Africa.

Interestingly, the results indicate a substantial contribution of boundary conditions (BCON) to the nitrate (NO3) component of PM2.5, which is averaged by country and calculated exclusively over land. The attached figure illustrates this finding. This is somewhat unexpected and challenging to interpret, given the anticipated predominant influence of local sources. It is noteworthy that this pattern is unique to NO3, whereas the source contributions for sulfate (SO4), ammonium (NH4), and secondary organic aerosols (SOA) appear more intuitively aligned with expectations.

In contrast, as shown in the below figure for SO4, despite a considerable contribution from BCON, there is a discernible impact from local sources, which is more pronounced when compared to NO3.

My preliminary hypothesis attributes this anomaly to the oxidation timescales of NOx, which are slower relative to those of SO2. However, I am eager to hear any thoughts or insights you might have,

any opinions would be appreciated.


Hi Nosha,

Your results are interesting. I would certainly expect very high boundary contribution within a small domain like the UK, but I don’t have experience with modeling in that part of the world to be able to suggest what is reasonable and what is not.

I can add that NOX certainly contributes to PM2.5 formation and that interaction is tracked by ISAM. That was one of the goals of developing this tool. Furthermore, species with longer lifetimes do transport further into the domain and then react with other precursors to create species such as nitrate.

In the case of sulfate, the source is basically only SO2 emissions, which has a lifetime of a bit over a week. Again, I don’t know for certain in your domain, but I would consider that shorter precursor lifetime would give local sources more chance for contribution competition.

You could try to to run some zero-out simulation to check local contributions. I guarantee that it will not match your results, but it might give more more information.