WRF-Chem: overpredicted PM concentrations
  1. To identify too high PM predictions in WRF-Chem, we should explore the vertical distribution of fire emissions used in the WRF-Chem run. Use these fire emission files: /data/lee/AUG2017/run/wrffirechemi_d01_2017-08-05_*
  2. In addition to the “regular” vertical plots, can you also compute the fraction of emissions assigned into each layer?
  3. To make sure if one of PM speciation dominates this over prediction, we will check each PM species.
Note that we are using “MOZART_MOSAIC_4BIN_AQ_KPP” chemical option, which uses the call sum_pm_mosaic_vbs4 and sum_vbs4 subroutines in “module_mosaic_driver.F” under /opt/apps/wrf/compiled/WRFV3/chem.  
Please take a look how PM2.5 is computed in “module_mosaic_driver.F” (see lines 1028 to 1061).  Can you plot individual PM2.5 speciations and get the fraction of them? You can use the basemap for this.
Please check it out this WRF-Chem post that is relevant to our high PM concentrations: https://groups.google.com/a/ucar.edu/forum/#!topic/wrf-chem/OEX-rdINk-I