README -- Describes contents of ExtData/CHEM_INPUTS/UCX_201409 folder 15 Apr 2019 GEOS-Chem Support Team geos-chem-support@g.harvard.edu Files: ====== Montreal_fix.nc -- File with mean surface volume mixing ratio of the following stratospherically-important species from years C2BR2F4 (aka H2402). CBR2F2, CCl4, CFC113, CFC114, CFC115, CFC11, CFC12, CH3Br, CH3CCl3, CH3Cl, CH4, H1211, H1301, HCFC141b, HCFC142b, HCFC22, N2O, Overview: ========= Seb Eastham wrote: H2402 is getting set to a different value in GEOS-Chem "Classic (aka GCC) compared to GCHP. Long story short: this is because, back when the ASCII input files (still used by GCHP) got converted to NetCDF (used by GCC), we had monthly data for all species EXCEPT H2402 (aka C2Br2F4), for which we had annual. It looks like, when I made the new files, I said that the H2402 monthly value should correspond to the nearest January (i.e. use 1990 for 1990-06-01, but 1991 for 1990-07-01) but this was inconsistent with the old formulation which said “just use the year number (i.e. the 1990 data for any day in 1990). The new netCDF file (Montreal_fix.nc) can be dropped in as a replacement for CHEM_INPUTS/UCX_201403/Init1D/Montreal.nc, and should work out of the box. Of course, I’d recommend instead that we make it be in directory CHEM_INPUTS/UCX_201904/Init1D/Montreal.nc and edit the code accordingly (two lines in ucx_mod.F). I certainly agree that in the long run these might be better set through HEMCO, but this will help sort out the discrepancies we’re seeing right now at least. I believe that the discrepancies seen in the other CFCs are due to small differences in vertical mixing -- these differences are all tiny.