This past week air modelers from across the country gathered in Raleigh, NC for the Air and Waste Management’s (AWMA’s) “Guideline on Air Quality Models: A New Chapter” conference. AWMA puts these on about once every two years or so, and they provide a great opportunity for modelers to tell stories, share ideas, and to hear from EPA about the latest developments in the modeling world. This meeting was no exception.
Early in the conference the modeling group at EPA took the podium to provide some updates on AERMOD. First and foremost, the much-anticipated new release of 40 CFR 51 Appendix W still isn’t quite here yet—they’re expecting release in November or December. Concurrent with that will be a new regulatory version of AERMOD and a finalized version of their new background concentration guidance. The changes to AERMOD probably won’t impact many of you and your modeling, and the background concentration guidance will include “real-world” examples of approaches to refining background concentrations. It’s the latter that I’m most interested in with this update.
EPA also spoke about modifications they’ve had to make in the last few months in response to data availability issues. AERSURFACE recently began running into problems with file resolution tolerance with data from the National Land Cover Database; that bug has been fixed.
Similarly, the long-time source for upper air data when running AERMET was recently taken offline. Fortunately, EPA had recently updated AERMET to accept upper air data in the Integrated Global Radiosonde Archive (IGRA) format so it was still possible to get the necessary data; however, issues were quickly uncovered with missing surface levels in the IGRA data that would cause AERMET to crash. EPA has instituted a fix for that as well.
For the past several years there has been a discussion in the modeling community as to whether continuing with AERMOD was the appropriate path—after all, AERMOD is roughly 30 years old and no longer state-of-the-art science. Of course, developing a new model from scratch is a colossal undertaking, so it wasn’t much of a surprise when EPA announced that they had met internally and decided that AERMOD will indeed not be replaced but rather they’re going to overhaul the Fortran code like they did AERMET to make it easier to update in the future and to improve runtimes. Personally I don’t think model runtimes are much of an issue with computing power these days, but I understand the reality of EPA not having the manpower/funding to create a new model.
Finally, you may be wondering about the picture that accompanies this post. Since I’d already shared on social media a picture of me at the Blue Sky Modeling tabletop display at this conference I thought I’d post a picture that illustrates why I didn’t get this blog up the day after the conference was over—within hours upon returning home last week I headed to northern Pennsylvania to hike the Black Forest trail in the Tiadaghton State Forest. That’s me at one of the scenic stream crossings. :-)