Aerosol thermodynamical equilibrium
The GEOS-5 met fields [reported the relative humidity (RH) field as fraction instead of percent.] Therefore:
Bob Yantosca (email@example.com) wrote:
- GEOS-Chem was expecting to read in RH in percent (0..100) but was actually reading in fraction (0..1). Then when we got to the places where RH was used in the code (e.g. ISORROPIA, drydep_mod.f, sulfate_mod.f), we were dividing that by 100 again. So then this would further reduce the maximum RH from order 1 to order 0.01.
- One of the symptoms that we saw was that HNO3 (in many locations, though not all) was completely wiped out in the PBL.
- Would you expect a reduction in HNO3 by ISORROPIA if it were given 100x too small RH? Do you know if ISORROPIA would give junk results in this case?
Havala Olson Taylor Pye (firstname.lastname@example.org) replied:
- The current version of ISORROPIA in the code definitely has problems at low RH and this is one of my primary motivations for updating it. I found that under certain conditions, at low RH, ISORROPIA would convert all nitrate (HNO3 + NIT) to the aerosol phase which leads to significant overpredictions in NIT (see attached plot if you're interested). In the attached plot, below an RH of about 0.30 (30%) all nitrate was converted to the aerosol phase (the total amounts of all the species and T are the same for each data point). So by multiplying the RH by 100 to correct it, you probably moved from a region where ISORROPIA makes no sense to a region where is actually gives good predictions. Hope this clears things up. I've been working on ISORROPIAII and it gives predictions very similar to RPMARES (as far as where nitrate concentrations are high and low) and very different from ISORROPIA currently in GEOS-Chem. Let me know if you want any more information.
--Bob Yantosca 13:27, 1 April 2008 (EDT)