Talk:New isoprene scheme prelim
RO2+HO2 Reaction Discussion
The difference plots show large absolute changes in isoprene, OH, NOx, MVK, MACR etc., but the relative changes for these species appear to be small. The largest relative effect of inclusion of the updated RO2+HO2 reaction rate occurs for species we expect to change, namely HO2 and H2O2. Does it make sense to proceed to the next step (i.e. including the Caltech chemistry mechanism)?
For Fabien's chemistry mechanism I will document the necessary changes and include them on the wiki for others to discuss and edit before including them in GEOS-Chem for benchmarking. Does this sound like a reasonable plan of action?
--Emarais 21:10, 5 December 2011 (EST)
Deposition for first generation isoprene nitrates
Fabien and I got similar results, that the first generation of isoprene nitrates is not sensitive to deposition, as their lifetime is dominated by reactions with OH and O3. For ISOPND, OH+ISOPND dominates. For ISOPNB, O3+ISOPNB dominates. This is based on the rates from Lockwood et al. (2010).
So there is no point to test the deposition rates for the first generation of isoprene nitrates. I think Fabien also made this point in his ACPD paper.
--Jmao 20:35, 6 December 2011 (EST)
I think we agreed that the chemistry should be updated more frequently than it has been in the past. With that in mind, using the isoprene scheme I wrote may not be the best strategy as it is difficult to modify/maintain.
I would suggest we start from the standard chemistry and add the key reactions that we think are important to include
1) IEPOX and RIP chemistry should be updated
2) we should allow for some recycling of NOx from isoprene nitrates. Ideally we would have a function that sets the recycling and that would be easy to modify in the future (from reaction with ozone and OH).
3) Isomerization. I am not sure how to implement it. The implementation I suggested was derived from previous studies where the isomerization rate was so fast that one could reasonably assume delta peroxy radicals were not important. With the experimental rate of Crounse et al., this assumption does not seem warranted. On the other hand, we do not want to calculate the distribution of peroxy radicals (i.e. include the fast exchange with O2). We should probably treat HPALD explicitly (as a tracer ?) and not assume it is readily photolyzed.
4) treatment of the beta/delta channels. This is related to the previous point. At the moment, in the high NOx case, we have either a frozen distribution (based on Paulot et al. 2009) or no delta at all (based on Peeters et al, 2009). This matters for the isoprene nitrate yield (more from the delta) and the fast production of hydroxyacetone/glycoladehyde/glyoxal/methylglyoxal. I am not sure how to include this while keeping with the overall goal of simplicity.
5) Update of ROOH+OH. That one is relatively straight forward.
--Fabien 20:35, 6 December 2011 (EST)