Non-linearities in reaction rates

[sahil.bhandari]

We discussed in the workshop how HYSPLIT only allows linear conversions from reactions. Could we run a separate box model (for the reactions), find an equivalent linear rate of conversion (including for non-linear systems) and then put that value in HYSPLIT? Is that a commonly accepted approach to address the non-linearity of actual reaction rates?

[christopher.loughner]

Yes, if you can generate an estimate of the real reaction mechanism and create an approximate first-order rate that gives a good representation of that overall rate, then I think you can try to do something in HYSPLIT. However, if you really want to do a lot of complex chemistry, it may be best to use an Eulerian chemistry model, like CMAQ, WRF-chem, or GEOS-chem.

Another thing to consider is if multiple emissions sources are impacting each other, then it would be necessary to run all of the emissions in HYSPLIT at the same time and grid the pollutants into an Eulerian grid to perform the chemistry each timestep. This has been done in the past with the CB4 chemical mechanism. It is possible to update the HYSPLIT model to use a more up to date chemical mechanism.

[sahil.bhandari]

Thanks so much!