CAMx: downloading input data from EPA OSAP 2022 modeling platform

Other Software
1

Describe your issue in detail

I am still new at using CAMx for real world applications, and am preparing a CAMx annual run using input files from the OSAP 2022 modeling platform on this AWS website: OSAP 2022 Modeling Platform - Registry of Open Data on AWS

I am hoping someone has done this before and can help guide me through the process. My questions are as follows:

  1. Are the input datasets from the website above in CAMx v7.32-ready format? Specifically

https://epa-2022-modeling-platform.s3.amazonaws.com/12US2_wrfcamx/

https://epa-2022-modeling-platform.s3.amazonaws.com/wrfcamx/

https://epa-2022-modeling-platform.s3.amazonaws.com/bcon/HEMI_CMAQ_12US2_CAMxBC/

https://epa-2022-modeling-platform.s3.amazonaws.com/bcon_v2/HEMI_CAMQ_12US2_CAMxBC_v2/

https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/CAMx/

https://epa-2022-modeling-platform.s3.amazonaws.com/MCIP/12US1/

  1. If any, which datasets are CAMx-ready?

  2. For the rest of the datasets, are there lists of input datasets that specifically require the use of one or more of the support software programs listed on this CAMx webpage?

What version of the other software are you using?

I am using CAMx v7.32 which supports both Fortran binary and NetCDF input files.

What steps have you taken to resolve this issue already?

I have compared these datasets to those from the CAMx test case and tried to match like for like, but it is unclear whether there is just one standard NetCDF data format or multiple formats based on the dataset contents. I have also tried Verdi to open some NetCDF files, but not all of the NetCDF files are readable using Verdi. Most can be viewed using ncdump, but that’s no guarantee that they are in the CAMx-ready format. There are a few datasets that I recognize as SMOKE output, and know that they require preprocessing.

This is the most complete collection of CAMx input data I have found, but it would be very helpful to know if there are alternative sources which require minimal CAMx preprocessing. I would really appreciate some guidance before I go ahead and download a year’s worth of these input files, given the limited disk space we presently have and the required preprocessing for other datasets.

I have only been subscribed to the CAMx mailing list for the last 2 years, so I don’t have access to any relevant discussions that occurred before then. Any assistance is much appreciated!

2

Hello all,

I managed to find my predecessor’s old CAMx runscript that used version 7.0 with input data from the 2016v1 EMP, which listed the actual filenames he downloaded for both the 36US3 and 12US2 grids. I am able to match most of them with files from AWS S3 Explorer , but the complete set of CAMx-ready NetCDF format files for ic/bc and emiss only seems to be to be available for the 12US1 grid from here:
ic/bc: https://epa-2022-modeling-platform.s3.amazonaws.com/wrfcamx/
landuse,met2d,met3d,snow_age,vdiff

emiss: https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/
beis_inline/
canmex_point/
cmv_c1c2_12/
cmv_c3_12/
merged_nobeis_norwc/
merged_withbeis_withrwc/
premerged_area/
afdust_adj/
airports/
afdust_adj/
airports/
beis4/
canada_afdust_adj/
canada_og2D/
canada_onroad/
canada_ptdust_adj/
canmex_ag/
canmex_area/
fertilizer/
livestock/
mexico_onroad/
nonpt/
nonroad_diesel/
nonroad_gas/
np_oilgas/
np_solvents/
onroad_ca_adj_diesel/
onroad_ca_adj_gas/
onroad_diesel/
onroad_gas/
openburn/
rail/
rwc/
pt_oilgas/
ptagfire/
ptegu/
ptfire-rx/
ptfire-wild/
ptfire_othna/
ptnonipm/
ptnonipm_hr/

Other than the ic/bc input files, there doesn’t seem to be specific assignments of emiss input files to the emiss_grid(?,?) arrays based on several old and new CAMx runscripts I’ve examined. I also noticed that some emiss input files for the v7.0/2016v1 EMP runscript are merged, such as ag (fertilizer+livestock?), fires (all fires), onroad (gas+diesel), nonroad (gas+diesel), onroad_ca (gas+diesel) and nonroad_ca (gas+diesel).

It seems feasible to start with a simple case where all emissions come from a single merged file such as those from merged_withbeis_withrwc/, since i don’t think CAMx explicitly models biogenics - I would like someone to confirm that these merged files contain all the separate emissions sources listed above and in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/. Also it may not be necessary to use the 36US3 grid if all input data is available for the 12US1 grid, and the 12US2 grid is a subdomain of the 12US1 grid anyway.

Thank you for any insights and direction you can provide!

3

Hi,

This is Joey from EPA, I am in charge of 2022 platform data release.

Both wrfcamx folders are CAMx ready files one is for 12US1 and the other one is for 12US2

We have two versions of 12US2 CAMxBC, you can find the detailed information for BCON version 1 in https://www.epa.gov/system/files/documents/2026-04/2022-v1-modeling-platform-aq-model-tsd_final_version.pdf

We are still working on v2 TSD, but the differences of BCON from hemi-CAMQ runs are:

Increased spin-up length

Correction to December formaldehyde emissions

Emissions files in your list are CAMx ready.

We testes all the files using CAMx v7.20, and you can find EPA 2022v1 base CAMx runscript in

Please let us know if you have any specific questions.
Thanks
Joey

4

For the 2022he CMAx model ready emissions these sectors were included

mrgpt: canmex_point, cmv_c1c2_12, cmv_c3_12, pt_oilgas, ptegu, ptnonipm, ptnonipm_hr, ptfire-rx, ptfire-wild, ptagfire, ptfire_othna

emis2d: canada_afdust_adj, canada_og2D, canada_onroad, canada_ptdust_adj, ocean_cl2, afdust_adj, airports, canmex_ag, livestock, nonroad_gas, nonroad_diesel, np_oilgas, np_solvents, openburn, onroad_gas, onroad_diesel, onroad_ca_adj_gas, onroad_ca_adj_diesel, canmex_area, rail, mexico_onroad, nonpt, fertilizer, beis4, rwc

These are considered complete emissions for the modeling platform.

The entire package for processing from the inventory to CMAQ and then converting to CAMx is available on the FTP. Note that this file is large (15 GB).

https://gaftp.epa.gov/Air/emismod/2022/v2/2022v2_platform_2022he_all_inclusive_SMOKE_package_07oct2025.tar.gz

A description of what is included in this package is also available.

https://gaftp.epa.gov/Air/emismod/2022/v2/2022v2_platform_2022he_all_inclusive_SMOKE_package_07oct2025_contents.txt

5

I can’t believe I missed this, but the list I was searching for of all 12US1 (and 12US2?) emiss sectors included in the merged file is in

https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/merged_withbeis_withrwc/README_2022he_cb6_22m_12US1_withbeis_withrwc_premerged_sectors.txt

But thank you @huang.jiaoyan, that is very helpful! Just to confirm that for NetCDF format inputs:
the 12US1 ic/bc is in https://epa-2022-modeling-platform.s3.amazonaws.com/wrfcamx/
the 12US1 emiss is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/premerged_area/ for individual sectors
the 12US1 emiss is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/merged_withbeis_withrwc/ for fully merged sectors
the 12US1 emiss/cmv_c3_12 is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/cmv_c3_12/
the 12US1 emiss/pt_oilgas is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/pt_oilgas/
the 12US1 emiss/ptnonipm is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/ptnonipm/
the 12US1 emiss/canmex_point is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/canmex_point/

the 12US2 ic/bc is in https://epa-2022-modeling-platform.s3.amazonaws.com/12US2_wrfcamx/
the 12US2 emiss is in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US2/cmaq_cb6ae7/merged_withbeis_withrwc/ for fully merged sectors

So if I was planning to simulate 12US1 pollutant emissions and transport during 2022 would I also need to add:
the 12US1 lightning_nox in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/lightning_nox/
the 12US1 ptagfire in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/ptagfire/
the 12US1 ptfire-rx in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/ptfire-rx/
the 12US1 ptfire-wild in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/ptfire-wild/
the 12US1 ptfire_othna in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/2022he_cb6_22m/12US1/cmaq_cb6ae7/ptfire_othna/

or for 12US2 pollutant emissions and transport during 2022 would I need to add:
the 12US2 ozone_col in https://epa-2022-modeling-platform.s3.amazonaws.com/CAMx_ancillary_inputs/ozone_col/2022_12US2_output/
the 12US2 photolysis rate in https://epa-2022-modeling-platform.s3.amazonaws.com/CAMx_ancillary_inputs/photolysis_rate/2022_12US2_output/

It seems other than the fully merged emiss sectors and ic/bc, 12US1 has data gaps for ancillary data while 12US2 has data gaps for emiss point sources.

Thank you for that list @james.beidler, I tried looking in different places but couldn’t find information about the contents of mrgpt (point?) and emis2d (area?). Consolidation would certainly make it easier to organize input data, but the 12US2 files in https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/CAMx/ are in *.camx and *.bin format which are won’t be supported in CAMx v7.4, so I’m hesitant to use those. For now I’m using CAMx v7.32, but I would prefer data in NetCDF format so I only have to download them once - a month at a time while I run a test series, and then the rest if all goes well.

6

Thank you for the feedback. In light of the change in CAMx v7.4 we are going to consider developing the CAMx emissions using netCDF rather than legacy UAM going forward.

7

Hello again,

I have identified the necessary 12US2 OSAP 2022 datasets from AWS S3 Explorer needed to run CAMx v7.32 thanks to @james.beidler and @huang.jiaoyan, as follows:

Initial_Conditions: https://epa-2022-modeling-platform.s3.amazonaws.com/bcon_v2/HEMI_CAMQ_12US2_CAMxBC_v2/ICON_2022he_cb6r5ae7_HEMI.v55.202???.camx
Boundary_Conditions: https://epa-2022-modeling-platform.s3.amazonaws.com/bcon_v2/HEMI_CAMQ_12US2_CAMxBC_v2/BCON_2022he_cb6r5ae7_HEMI.v55.202???.camx

Surface_Grid: https://epa-2022-modeling-platform.s3.amazonaws.com/12US2_wrfcamx/landuse.2???.12US2.zhang.camx.nc.salt
Met3D_Grid: https://epa-2022-modeling-platform.s3.amazonaws.com/12US2_wrfcamx/met3d.2???.12US2.35.wrf.camx.nc
Met2D_Grid: https://epa-2022-modeling-platform.s3.amazonaws.com/12US2_wrfcamx/met2d.2???.12US2.35.wrf.camx.nc
Vdiff_Grid: https://epa-2022-modeling-platform.s3.amazonaws.com/12US2_wrfcamx/vdif.2???.12US2.35.wrf.camx.0.1.nc.YSU.kvpatch

Emiss_Grid: https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/CAMx/emis2d.202???.12US2.2022he_cb6_22m%2Bss.camx
From @james.beidler, area sources including:
canada_afdust_adj, canada_og2D, canada_onroad, canada_ptdust_adj, ocean_cl2, afdust_adj, airports, canmex_ag, livestock, nonroad_gas, nonroad_diesel, np_oilgas, np_solvents, openburn, onroad_gas, onroad_diesel, onroad_ca_adj_gas, onroad_ca_adj_diesel, canmex_area, rail, mexico_onroad, nonpt, fertilizer, beis4, rwc

Point_Sources(1): https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/CAMx/mrgpt.202???.12US2.2022he_cb6_22m.camx
From @james.beidler, merged point sources including:
canmex_point/
cmv_c1c2_12/
cmv_c3_12/
pt_oilgas/
ptagfire/
ptegu/
ptfire-rx/
ptfire-wild/
ptfire_othna/
ptnonipm/
ptnonipm_hr/

Point_Sources(2): https://epa-2022-modeling-platform.s3.amazonaws.com/emis/2022v2/CAMx/ptsr.lnox.202???.12US2.bin
Point sources for lightning NOx

According to @huang.jiaoyan in a previous post in this thread, the OSAP 2022 CAMx datasets were run for version 7.2. When I used the Chemistry_Parameters, Ozone_Column and Photolysis_Rate downloaded from the OSAP 2022 data bucket for CAMx version 7.32 this triggered error messages like so:

Using Chemistry_Parameters = CAMx7.2.chemparam.CB6r5_CF2E

CAMx version in CHEMPARAM file is INVALID
Expecting: VERSION7.3
Found: VERSION7.2

ERROR in READCHM.

Using Chemistry_Parameters = CAMx7.32.chemparam.CB6r5 with Photolysis_Rate = rate.20211222.12US2.do.CB6r5.txt

ERROR in READPHT:

ERROR: Reading photolysis rates
Reading header record at line: 1

Photolysis label record is INVALID
Expecting: TUV4.8CAMx7.30_CB6r5
Found: TUV4.8CAMx7.20_CB6r5
Make sure you are using the correct version of the TUV processor for this
version of CAMx and chemistry mechanism.

Thanks to @huang.jiaoyan, new sets of Chemistry_Parameters, Ozone_Column and Photolysis_Rate that work in CAMx version 7.32 have been provided here:

Chemistry_Parameters: https://epa-2022-modeling-platform.s3.amazonaws.com/Model_jobs_v2/2022he_CAMx_base_12US2/CAMx7.3.chemparam.CB7r1_CF3E_COMPLX_NH3Rscale0
Ozone_Column: https://epa-2022-modeling-platform.s3.amazonaws.com/CAMxv7.31_ancillary_inputs/ozone_col/ozcol.202???.12US2.txt
Photolysis_Rate: https://epa-2022-modeling-platform.s3.amazonaws.com/CAMxv7.31_ancillary_inputs/photolysis_rate/rate.202???.12US2.do.CB7r1.txt

One dataset I was unable to find in the OSAP 2022 bucket was Cloud_Grid, but it seems CAMx version 7.32 can run without it.

If others are working with 12US2 grids like I am, the changes that one needs to make in /CAMx7v32/src.v7.32/Includes/camx.prm are:
change parameter ( MXCELLS = 200 ) to parameter ( MXCELLS = 396 )
change parameter ( MXPTSRC = 700000 ) to parameter ( MXPTSRC = 914784 )

One question for more experienced modelers, I noticed that the OSAP 2022 datasets have a mix of different chemical mechanisms like CB6 and CB6r5ae7, with the new ones from @huang.jiaoyan using CB7r1. If I repeat my test run using CAMx version 7.2 as EPA did, would these different mechanisms cause enough deviation(s) so that my results won’t be comparable to theirs in https://epa-2022-modeling-platform.s3.amazonaws.com/CAMx_output_v2/2022001.12US2.35.2022he_cb6_HEMI_NH3Rscale0v2_AMD.epa.camx.avrg.grd01.nc for example?

Many many thanks to @james.beidler and @huang.jiaoyan!

Note: the question marks in the URLs are wildcard characters, the forum editor will only let me use 3 when there are actually 5 in each YearMonthDay section of the filename