Community Modeling and Analysis System


(Sparse Matrix Operator Kernel Emissions) Modeling System


SMOKE version 4.9 Release Notes

Starting with SMOKE v4.9, we are working to make the SMOKE distribution packages and test data more compatible with the directory structures and run scripts used in EPA's Emissions Modeling Platforms.

If you are interested in using SMOKE v4.9 with the older SMOKE Assigns files and run scripts, we recommend downloading the SMOKE v4.8.1 packages and replacing the executables with those from SMOKE v4.9.

BEIS4 and BELD6 Support

  • New programs Normbeis4 and Tmpbeis4 support using BEIS4 emissions factors and BELD6 land use data

Smkreport Updates

  • Column headers in point source reports now match Flat File 2010 terminology (Facility ID, Unit ID, Rel Point ID, Process ID)
  • New "BY FACILITY" option in the REPCONFIG file replicates the existing "BY PLANT" instruction
  • New "BY BOILER" REPCONFIG option summarizes emissions by ORIS Boiler ID

Bug Fixes

  • Elevpoint: fixed error that caused uniform stack flow (STKFLW) values to be output to the STACK_GROUPS file
  • Gentpro: fixed output of county codes in NetCDF files
  • Movesmrg / Smkmerge: Fixed errors when using subsector grouping
  • See RELEASE NOTES above for a complete list of updates

This is the official SMOKE website hosted by the Center for Environmental Modeling for Policy Development (CEMPD) at the University of North Carolina at Chapel Hill. SMOKE is an active open-source development project supported and distributed by the CEMPD through the Community Modeling and Analysis System Center.

SMOKE Online Training


Previous Documentation

Online Resources


System Requirements

O/S Linux
Processor 64-bit x86
Memory >2 Gb RAM
Disk Space > 40 Gb
Software I/O API, netCDF, Fortran ( Portland Group | Intel )

SMOKE-ready Data

SMOKE input data consist of emissions inventories, temporal and chemical speciation profiles, spatial surrogates, gridded meteorology and land use data, and other ancillary files for specifying the timing, location, and chemical nature of emissions. SMOKE is distributed with example data for getting started with the model. The example files distributed with SMOKE are for demonstration purposes only, they are not meant for real-world modeling applications.

The primary source for non-meteorology SMOKE input data is the U.S. EPA Clearinghouse for Inventories and Emissions Factors (CHIEF). The U.S. EPA Office of Air Quality Planning and Standards (OAQPS) Emissions Inventory and Analysis Group (EIAG) provides SMOKE inputs for different rule-making modeling platforms. These platforms include not only the NEI for both criteria air pollutants (CAPs) and hazardous air pollutants (HAPs), but also all of the SMOKE ancillary data files created by EPA for use in SMOKE. EPA uses CHIEF to provide these data.

Meteorology data must be generated for specific SMOKE applications using either MM5, WRF, or a similar model. The output data from meteorology models must be formatted for SMOKE using a program like MCIP.

History of SMOKE

The Sparse Matrix Operator Kernel Emissions (SMOKE) Modeling System was originally developed at MCNC to integrate emissions data processing with high-performance computing (HPC) sparse-matrix algorithms. SMOKE is now under active development at the Institute for the Environment and is partially supported by the Community Modeling and Analysis System (CMAS).

SMOKE is primarily an emissions processing system designed to create gridded, speciated, hourly emissions for input into a variety of air quality models such as CMAQ, REMSAD, CAMX and UAM. SMOKE supports area, biogenic, mobile (both onroad and nonroad), and point source emissions processing for criteria, particulate, and toxic pollutants. For biogenic emissions modeling, SMOKE uses the Biogenic Emission Inventory System. SMOKE is also integrated with the on-road emissions model MOVES.

The sparse matrix approach used throughout SMOKE permits rapid and flexible processing of emissions data. Rapid processing is possible because SMOKE uses a series of matrix calculations rather than a less-efficient sequential approach used by previous systems. Flexible processing comes from splitting the processing steps of inventory growth, controls, chemical speciation, temporal allocation, and spatial allocation into independent steps whenever possible. The results from these steps are merged together in the final stage of processing using vector-matrix multiplication. This means that individual steps (such as adding a new control strategy, or processing for a different grid) can be performed and merged without having to redo all of the other processing steps.

SMOKE is written in Fortran 90 and is designed to run on a variety of UNIX platforms. We currently provide executables for Linux and the source code is available for download and can easily be compiled for your particular system. We do not support running SMOKE on Windows, due to the inherent limitations of that system. The current version of SMOKE is version 4.9, although earlier versions are still available for download.