For the last decade, EPA has been revising, updating, fixing bugs, and introducing beta options in the AERMOD model with the goal of providing a dispersion modeling tool that is dependable, accurate, and reflective of the best current science. Dispersion models are a tool that play a key role in both the permitting process and air management for nonattainment areas. Therefore, it is imperative that their output provide an accurate picture of atmospheric concentrations associated with industrial and urban area sources of emissions.

As part of its mandate, EPA's Office of Inspector General (OIG) conducted an internal audit to assess the effectiveness of the process for reviewing and approving air quality dispersion models. This review and its findings are consistent with supporting EPA's mission of providing the best tools for estimating air quality impacts associated with emissions from industrial, mobile, and commercial sources. Emphasis in this review by the OIG was in the Guideline on Air Quality Models, Appendix W of 40 CFR Part 51 and on the American Meteorological Society/EPA Regulatory Model (AERMOD) which is EPA's preferred model for most regulatory modeling.

OIG Audit Findings

In its September 2018 report, the OIG found, in part, that from 2006-2016, EPA made changes to AERMOD resulting in 117 bug fixes, 56 enhancements, and 65 miscellaneous changes. These updates, changes, and additions were included in a total of 12 updated versions of AERMOD (the most current version is 18081 from the Spring of 2018). These frequent revisions to the model make permitting and state implementation plan (SIP) development difficult. Model updates and new versions, produce changing results however few model performance evaluation updates have been forthcoming to assist local, state, regional, and national modelers in determining how to interpret these changes. The OIG's audit covered EPA's Office of Air Quality Planning and Standards (OAQPS) Air Quality Modeling Group (AQMG) model review, evaluation, and modification procedures. The findings indicated needed changes to the process for reviewing new and revised models. After reviewing the Guideline on Air Quality Models, model Clearinghouse memoranda, user's guides, model formulation and evaluation studies, and other EPA reports, the OIG report concluded that guidance was available for new models but not available, nor followed, for model revisions. Although the focus of the OIG report was AERMOD, processes for other models were assumed to be similar.

Based on this audit, the OIG made four recommendations as required corrective actions to bring EPA guidance for updating models into compliance with government specifications.

  • Develop Standard Operating Procedures (SOP) to guide and document the process for reviewing and approving revisions to EPA recommended models.
  • Develop Quality Assurance Program Plans (QAPP) to describe the results of systematic planning, updating, and evaluation of model revisions.
  • Update the OAQPS Quality Management Plan to include the SOP and QAPP documents.
  • Train OAQPS staff to consistently follow the SOPs and QAPPs.


The EPA modeling group and OAQPS have responded to all of these recommendations in a positive, proactive manner. One of the design goals for AERMOD was an ability to “accommodate modifications with ease as the science evolves.” Another goal was to incorporate new science. Combining these goals, EPA began work in 2017 on an AERMOD Development Plan which fulfills portions of the first two recommendations from the OIG audit. The initial elements of this plan were outlined in a series of white papers released by EPA on September 19, 2017. While the white papers emphasized specific science topic areas that were being considered for AERMOD updates, they were intended to publicly alert the modeling community to areas of EPA focus as it developed the overall Modeling System Development Plan. For AERMOD, these areas include low wind options, moisture saturated plumes, improved downwash algorithms, and updated NOx to NO2 conversion techniques. Two other areas addressed in the white papers include an improved roadway modeling tool called R-Line which could be incorporated into AERMOD, and an improved overwater model that also could be added to AERMOD.

Implications for Industrial Sources

As a result of the OIG audit and subsequent EPA process changes, the AERMOD revision process will be more rigorous than in the past including the use of the SOP and QAPP. These documents will include thoughtful programming elements and will improve EPA's research, use, and release of revisions to the work-horse of air permit modeling, AERMOD.

Part of the enhanced process has already been established and is being used in the current version of AERMOD, Version 18081. Instead of new options being introduced in a beta format, a two-layer process of alpha (for research) and beta (for testing in a regulatory setting) versions for proposed enhancements and changes has been introduced. As EPA follows the new QAPP, changes may come a bit more slowly, allowing a permit applicant to proceed with the knowledge that a proposed model in a modeling protocol or air permit application will still be acceptable when the modeling is complete.

The AQMG has responded positively to all of OIG's recommendations and the status of EPA responses is summarized on page 15 of the OIG report. The OIG has accepted the current and planned corrective actions which are summarized in a letter from EPA Assistant Administrator of EPA, William Wehrum and included as Appendix A in the OIG report.