PM underestimation in the EMEP model
Participants: DS,ST,JEJ,BMS,ACB,MG,HF,PW,MS
Possible reasons:
* pH dependency of SO2 ox rate
* wet scavenging (in ice, scavenging rates …)
* emission heights:
DS:I have started this, testing different plume-rise
formula together with the IIASA . Peter contributed with a subroutine to
spread the emissions into flexible vertical layers also. Should all be
ready soon-ish, but probably won't change things very much as you say. (probably not important)
* Size of vertical layers, boundary layer mixing
* Secondary Organic Aerosol contribution
* Agricultural and Industrial dust in windy, dry conditions
* BIC for aerosols, dust in southern Europe
* Wood burning emissions underestimated:Robert is doing a lot with Hugo as we speak, but the problem is with the “official” emissions. We can't change them, even
when we know they are wrong.
We could change the time-series though, using a “degree-day” approach as
Betrand used. That would be good, but requires some testing. I think
that SNAP-2 in general should have this, and wood-burning especially,
but I also suspect that wood-burning starts at a lower temperature than
Betrand's 20C degree days.
* Station representativity for some specific sites (Susceptibility to inversion situations in winter?)
* Kz, Hmix - currently our minimum mixing height is in the 2nd layer
(k=19). I have some test versions which allow very shallow mixing
heights. These cause a slight increase in surface PM, and are
scientifically ok I think. I can commit next week if all looks well.
(What I am less sure about is what we should do for Kz above Hmix. Now
we have zero, earlier we used critical Richardson numbers, etc. Seems
better to use the latter.)
* I suggested some ideas for David H. here to work with, but that seems
to be progressing slowly. I'll check again on this next week.
* Ammonium nitrate formation
* Emission seasonality for SO2 missing
Which Analysis :
* Separate discussion by aerosol component
* AMS measurements EUSAAR ?
* Caliop extinction profiles for European subregions become available shortly, extinction profile in model needed,
* EUCAARI aircraft measurements?
* Seasonal and spatial bias analysis
** When do we underestimate?? It would be good to use time to figure out
if we can see patterns in the days when we underestimate compared to
when we don't. Is it on very cold days, or generally for example? Which
compounds cause most of the problem? Might give a clue on whether it is
meteorology of chemistry.