I am running CMAQv5.4 (saprc07tc) with ISAM for a few months (May-June-July) around Southern California at 2km resolution and am getting strange ozone results. Namely, ISAM is telling me that even more than 2 months in, initial conditions are contributing 5-10ppb of ozone (roughly 10%). That seems outrageously large for that far away from the start date. The ICON contribution is also not monotonic, and increases, for example, between late June and early July.
I don’t understand how this is physically possible so I’m inclined to think it’s a setup issue or some obscure bug, but I welcome any ideas.
The model was compiled with ifort 14.0.3. I’ve attached my run script, ISAM control file, and images of the mean ozone on each of the days I mentioned. The ICON contribution issue seems to hold across the entire run, and looks similar for another run I did for August-September with similar settings.
Thanks for the reply. We’re using a custom build of v5.4 that has an updated SAPRC mechanism, so switching versions isn’t a small undertaking. Perhaps at some point we might either port up our SAPRC mechanism or port down the ISAM module changes. If you have recommendations for anything I could test configuring within our existing build, I’m certainly open to ideas. If there’s no good answer for now other than “this is a bug that has since (probably) been fixed”, then I’ll live with that too.
How were your IC/BC inputs to CMAQ created? Were they processed from a global model?
One suggestion from my colleague is to use clean BC/IC’s.
Try running ICON and BCON using the profile option.
This may help rule out any issue with the IC/BCs.
