11. Integrated Source Apportionment Method (CMAQ-ISAM)
The Integrated Source Apportionment Method (ISAM) calculates source attribution information for user specified ozone and particulate matter precursors within the CMAQ model. CMAQ-ISAM has been substantially updated in the CMAQv5.3 release.
The CMAQ model provides users the concentration and deposition fields of many pollutant species. These species are usually combinations of different types of primary emissions and secondary formation that have been physically and chemically transformed in the model. However, sometimes it is desirable to know specific source attribution information for the model outputs. For example, how much of the ozone in an urban area was formed due to nitrogen oxides emitted from motor vehicles in a neighboring state?
Answering this type of question often requires running an air quality model twice, once with the standard emissions scenario and once with the source of interest completely removed. The difference between the two runs is then assumed to be attributed to the removed source. While this approach is reasonably straightforward to implement, it has some drawbacks. For example, removing a large source from the system in a highly nonlinear chemical mixture can lead to some errors. Also, calculating source attribution of many sources can be logistically and computationally prohibitive.
Alternatively, running CMAQ with ISAM enabled allows the user the ability to calculate source attribution of a large number of sources directly by the model in one simulation.
Note: While full model species list apportionment is in development, currently ISAM is limited to the following species classes in CMAQ:
SULFATE - ASO4J, ASO4I, SO2, SULF, SULRXN NITRATE - ANO3J, ANO3I, HNO3, ANO3J, ANO3I, HNO3, NO, NO2, NO3, HONO, N2O5, PNA, PAN, PANX, NTR1, NTR2, INTR AMMONIUM - ANH4J, ANH4I, NH3 EC - AECJ, AECI OC - ALVPO1I, ALVPO1J, ASVPO1I, ASVPO1J, ASVPO2I, ASVPO2J, ASVPO3J, AIVPO1J, VLVPO1, VSVPO1, VSVPO2, VSVO3, VIVPO1 VOC - 21 species in CB6R3 (see OZ_DEFN.F) PM25_IONS - ACLI/J,ANAI/J,AMGJ,AKJ,ACAJ,AFEJ,AALJ,ASIJ,ATIJ,AMNJ,AOTHRI/J CO - CO OZONE - all NITRATE species + all VOC species
11.2 Build Instructions
Starting with CMAQv5.3 model release, ISAM is provided directly with the source code of the base model. To use ISAM, follow the normal build process for CMAQ described in Chapter 5 but make sure to uncomment the following line in bldit_cctm.csh:
A note about I/O API installation for ISAM applications
I/O APIv3.2 supports up to MXFILE3=64 open files, each with up to MXVARS3=2048. ISAM applications configured to calculate source attribution of a large number of sources may exceed this upper limit of model variables, leading to a model crash. To avoid this issue, users may use I/O API version 3.2 "large" that increases MXFILE3 to 512 and MXVARS3 to 16384. This version is available as a zip file from the following address:
Installation instructions for I/O API v5.3-large are provided in README.txt in the .tar.gz file.
11.3 Run Instructions
To begin a CMAQ simulation with source apportionment enabled, the ISAM section of the runscript must be configured. The additional necessary environment variables are listed in Table 11-1.
Table 11-1. ISAM run script variables
|CTM_ISAM||Y/N||Set this to Y to enable ISAM|
|SA_IOLIST||path/filename||Provide the location of the ISAM control file (discussed below)|
|ISAM_BLEV_ELEV||" MINVALUE MAX VALUE "||LAYER range for the instantaneous ISAM output concentrations|
|AISAM_BLEV_ELEV||" MINVALUE MAX VALUE "||LAYER range for the average ISAM output concentrations|
|ISAM_NEW_START||Y/N||set Y for a new simulation and N for continuing from a previous day's outputs|
|ISAM_PREVDAY||path/filename||Provide the location of the previous day's ISAM restart file|
|SA_ACONC_1||path/filename||ISAM output for average apportioned concentrations|
|SA_CONC_1||path/filename||ISAM output for instanteneous apportioned concentrations|
|SA_DD_1||path/filename||ISAM output for apportioned dry deposition|
|SA_WD_1||path/filename||ISAM output for apportioned wet deposition|
|SA_CGRID_1||path/filename||ISAM output for a restart file to continue the simulation further in time|
Additionally, ISAM can track emissions confined to geographic regions. This functionality can be enabled through CMAQ's
RegionsRegistry set in the
EmissCtrl namelist (Appendix B.4) and is discussed further below.
11.3.1 ISAM control file (SA_IOLIST)
SA_IOLIST is a text file used to configure which tag classes, emissions streams, and source regions the model will track. An example of this file,
isam_control.txt, is provided in $CMAQ_HOME/CCTM/scripts. The formating of this file must be kept intact, but it does allow for insertion of comment lines.
Each ISAM simulation requires the specification of the
TAG CLASSES that the user desires to apportion. The current list includes the following choices
SULFATE, NITRATE, AMMONIUM, EC, OC, VOC, PM25_IONS, CO, OZONE. Species associated with each of these are provided in section 11.1. One or more of these tag classes must be specified in
SA_IOLIST. Multiple tag classes are comma delimited.
TAG CLASSES |OZONE, SULFATE
After setting tag classes for the simulation, information for one or more tags is required. Each individual tag will track the species from the specified
TAG CLASSES and has its own set of three options in the control file. The first option is the name:
TAG NAME |EGU
It is recommended that the text string for the tag name be kept short (ideally three characters) in order to accommodate the longer species names from some chemical mechanisms in the ISAM output files.
The second option is the comma delimited list of regions to track with this tag. The keyword 'EVERYWHERE' is used to track domain-wide emissions. To track region-constrained emissions, variable names from the regions file specified in the
EmissCtrl namelist are used instead of the "EVERYWHERE' keyword. The regions file requirements are identical to the optional file used to scale emissions in predetermined geographical areas (Appendix B.4).
REGION(S) |NC, SC, GA
Finally, the emissions streams labels are required as the third option in the control file. These are the labels set in the runscript for the base CMAQ simulation.
EMIS STREAM(S) |PT_EGU, PT_NONEGU
The final line in the control file needs to be kept unchanged in order to aid the file parser in reading this file.
In addition to the user-specified list, ISAM will alway track and output three additional default tags with every simulation (note, that at least one user-specified tag must be defined):
ICON - contribution from initial conditions specified for the first day of the simulation BCON - contribution form boundary condtions throughout the simulation OTHR - contribution from all emissions not tagged by the user through isam control file.
11.4 ISAM Benchmark data
The input files for the CMAQv5.3 ISAM benchmark case are the same as the benchmark inputs for the base model, described in the CMAQ Benchmark Tutorial. Output source apportionment files associated with the sample
isam_control.txt provided in this release package are included in the benchmark outputs for the base model.
The CMAQ benchmark data can be downloaded from the CMAS Center Data Warehouse SE53BENCH Google Drive folder. The CMAQ benchmark test case is a two day simulation for July 1-2 2016 on a 100 column x 80 row x 35 layer 12-km resolution domain over the southeast U.S. Input and output files for a two week case covering July 1-14, 2016 are also available within the same Google Drive folder.
- Metadata for the CMAQ benchmark inputs: https://doi.org/10.15139/S3/IQVABD
- Metadata for the CMAQ benchmark outputs: https://doi.org/10.15139/S3/PDE4SS
Kwok, R.H.F, Napelenok, S.L., & Baker, K.R. (2013). Implementation and evaluation of PM2.5 source contribution analysis in a photochemical model. Atmospheric Environment, 80, 398–407 doi:10.1016/j.atmosenv.2013.08.017.
Kwok, R.H.F, Baker, K.R., Napelenok, S.L., & Tonnesen, G.S. (2015). Photochemical grid model implementation of VOC, NOx, and O3 source apportionment. Geosci. Model Dev., 8, 99-114. doi:10.5194/gmd-8-99-2015.
Sergey L. Napelenok, Computational Exposure Division, U.S. EPA