Last Thursday night, February 20, 2025, the CEO of Boulder A.I.R. (www.bouldair.com) Dr. Detlev Helmig gave a compassionate, articulate and, above all, useful report on our air quality trends in the Front Range over the last few years. It was incredibly reassuring to hear and watch an expert interpret complex air quality data into graphs and graphics that were completely understandable. I’ve never seen any such comprehensive approach by any of our state air quality agencies; and, for my money, they could learn a few things about doing so from this presentation.

The meeting was hosted by 350Colorado (350colorado.org), and directed by Bobbie Mooney, the staff attorney and an original founding member of the organization. There were around a couple of dozen attendees who had lots of questions at the end of Dr. Helmig’s talk, and he graciously stayed on to continue answering their questions well after the allotted hour was passed.

The recording of the presentation is available here; and the slide deck is available here.

I did grab a few screenshots which will give you an idea of the high level of scientific skill and integrity involved, and how much better informed we the public could be about our air quality (if our state air regulatory agencies really cared to do so). This would go a long way towards building trust in the data we were shown; and trust is in rather short supply with our air quality regulatory agencies (more on that later).

In the screenshots that follow, some of them contain a lot of dense information, and these small scale images are insufficient to fully convey what was shown; but they are linked to larger image, so click on the images in the post to see the larger images.

Please refer to the recording when it becomes available for a better understanding of them. (I can also make the image files available upon request, with the permission of 350Colorado.)

Screenshot 1

This first shot was an overview of the battlefield (so to speak), where the red dots are active wells, the blue dots are permits under consideration, and the big orange X’s are the six (now five) Boulder AIR monitoring stations, which are in Erie, two in Longmont, and two in Broomfield. Sadly, the first Boulder AIR monitoring station, which had been operating at the Boulder Reservoir since 2017, ceased operation on December 31, 2024 because the Boulder County Commissioners decided it was “no longer necessary” — which many disagree with. Dr. Helmig expressed regret over that decision, since that station had the longest data record, and therefore the greatest statistical significance.

Screenshot 2

This next shot shows where the areas of ozone non-attainment are across the US. (Non-attainment means “unsafe”, as in not attaining safe levels of ozone during ozone season, typically June to August, though there can be exceptions.) The red dotted circle is the nine county area from north Denver to the Wyoming state line, which is classified as ‘serious/severe’ non-attainment, shaded purple. The only level above that is ‘extreme’, which can be seen covering the Central Valley in California. This shows the comprehensive approach by Dr. Helmig, who has been researching surface level ozone for over 30 years, and is the author of over 200 articles in scientific journals. This is the kind of expertise that is reassuring to hear from when it comes to our local environmental quality.

Screenshot 3

The next slide shows data about particulate intensity during the wildfires we had over 2020 to 2023, which included the largest wildfires ever seen in Colorado or Canada. Note the enormous swings in the graphs; but what does mean practically?

Screenshot 4

To answer that question, see the photograph on the right, taken at the Union Reservoir, I believe, on May 19, 2023, where the limited visibility is striking….I can just imagine how the air must have smelled!

Screenshot 5

This next slide is from a recent (2025) paper published in Elementa:  Science of the Anthropocene, a scientific journal about climate change, where the emphasis is on pragmatic measurement of human impact on the natural environment due to climate change effects. Such articles require a high level of training just to understand them, but that wasn’t the point of the slide; the point was to show how the state regulatory agencies in Colorado (think CDPHE) mislead the public, perhaps not even intentionally, but their misuse of their own data. I think an entire presentation could likely be given on that topic alone!

Screenshot 6

Finally, this last screenshot shows how Boulder AIR data can be combined with meteorological data (wind speed, direction and height) to create very interesting color-coded two dimensional maps showing estimated pollution intensity over an area over a given period of time. This is a powerful analytical tool, as it can map with great precision where pollution originates. These maps are created by data, not by any artistic interpretation.

This one is of  interest because it covers an “emissions event” of acetylene from a drilling site in Broomfield during Aug-Oct 2020. I include it just to show the potential of how Boulder AIR data can be processed and analyzed — when there is focused goal for doing so.

Why Aren’t Boulder AIR stations more used?

While these screenshots show the powerful analytical and graphical capabilities of Boulder AIR to measure the pollutants in our air with great precision, it does not explain why these powerful instruments have not been adopted by our existing regulatory agencies.

The main reason is cost. Boulder AIR stations are expensive to set up, and expensive to maintain. The equipment is highly sophisticated (a gas chromatograph and mass spectrometer form the heart of them), it takes highly skilled personnel to set up, calibrate and maintain them, and the equipment itself must be housed in a climate controlled environment. And Boulder AIR is the only company that offer the skill, work experience and local presence for this exceptional tool.

A second reason is the flat refusal of our state regulatory agencies to accept Boulder AIR data as valid. The Larimer Alliance refuses to accept this, and will continue to challenge our regulatory agencies (i.e. CDPHE & APCD) on this point: on what basis can they refuse this data as valid? Boulder AIR has high standards (because their CEO is a published air chemistry researcher in many academic journals), and their data has been validated by the EPA, and has been accepted into their data repository for remote sites for years.  This question of the refusal of the state to accept Boulder AIR data begs a larger question: does the CDPHE suffer from regulatory capture? Which is to say: does our primary regulatory agency, charged with protecting our public health at large, actually suffer from regulatory capture, and is therefore more interested in protecting the very industry they are supposed to be regulating? That is a big question which the Larimer Alliance will  continue to investigate.

So why did the local jurisdictions make such an expensive investment? I would suggest the following reasons:

• there has been a severe ozone air pollution problem in the Front Range since the early 2000’s
• there has been nearly zero progress on its reduction by state regulatory agencies during this time
• the state has not provided or developed any comprehensive plan for dealing with the problem
• the state has not provided enough information to keep the public informed about why air quality is so bad
• local jurisdictions wanted to know exactly what was in their air, particularly areas affected by oil and gas operations, nearly always fracked wells, which had continued to encroach on residential areas in these jurisdictions, despite the pollution it causes

The only thing that the CDPHE has done in response is to develop an air quality indicator tool that informs the public when air quality is poor (big whoop!) The need to fill this information void created by the lack of data and transparency from the state would seem to be the reasons for this costly investment. Because Boulder AIR data from the various sites are published within about 15 minutes of sampling, they serve as real time monitoring stations for those communities. (Of course, the interpretation of that scientific data, and alerts to the public, could always be improved.)

Has the investment been worth it? That’s a complex question that only the administrators and benefactors of the data who live in these areas can answer. It is an ongoing process, which unfortunately has resulted in public confusion due to the politicalization of the issue. As anyone who has followed the issue of O&G regulation since the Polis administration took office, much of the blame for this lack of progress and public confusion can be laid at the feet of this administration; but that is another issue to be discussed at another time.

One of the questions that arises is are there less expensive ways to monitor our air quality? That depends on the technology used…

A Comparison of Boulder AIR and “canister-based” technology

A competing air monitoring technology that has been used in local jurisdictions is canister-based technology, which costs less in initial capital cost and later operation, is easier to set up and maintain, and is also available from a local company, Ajax Analytics. (see linkedin.com/company/ajax-analytics)

However, this technology operates quite differently, which deserves some examination.  The public deserves to know enough how to make their own decisions about these competing technologies in case their local elected leaders solicit public input from their constituents.

Canister technology, as the name implies, relies on an empty metal container — they can vary in size from a water bottle to a basketball:

— which captures an air sample. The rate at which the air sample is gathered can vary, from all at once to more slowly over time. When the sample is taken is another critical part of the design, which is trigged by a sensor attached to the canister. This sensor is suppose to detect the target chemical(s) (i.e. air pollutants), meaning it “knows” when to start taking air samples; this, too, has been criticized. Obviously, the canister has to stop its air sampling when it fills up. The canister can be configured regarding the size of the air sample it takes; it can fill up the entire canister at once, or it can take multiple samples over time until it is full. Once full, the canister is manually retrieved from its monitoring site and taken to a lab, where its air sample is analyzed; after which, the canister can be reused.

The weakness of this design are 1) the limited size of the sample, 2) the discontinuous timing of samples, due to the need for physical retrieval, 3) the cost of the manual transport of the canister from field site to a lab, and 4) the possible unreliability of the sensor triggering device.

This last reason is a real factor, because the sensors that trigger canisters have been known to get falsely triggered (such as rain), known as false positives. This is why the PID type sensors are not used in the scientific community, are typically used in handheld sensors. (see the photoionization detector wikipedia entry)

If the sensor is triggered by a false positive, the entire sample is obviously worthless. This results in a higher labor cost, and the delay in time, due to the need to transport canisters to and from a lab for analysis, since the false sample is not known until it is analyzed. So, although there is a lower capital cost for canister-based monitors initially, the important statistic is the cost per sample, which is a more accurate basis on which to compare these two types of systems.

I do not know what the cost of canister-based systems are, since that is proprietary to Ajax Analytics and its customers, but it is not hard to see that an automated system like Boulder AIR, which is taking continuous samples, once a minute, across a suite of thirteen target pollutants, will produce 18,720 data samples in one day, or 6,589,440 samples over a year. A canister system creates a sample only when it is triggered; and must then be physically picked up and analyzed. Such a system will tell you nothing about the regional air quality; it is only tell you something about the pollution concentration, assuming it is accurately triggered, and assuming the wind is blowing towards the canister to accurately capture a representative sample. In other words, there is hardly any comparison between a canister system and a continuous system. They are built for two very different purposes: one is for on-site emission event detection, the other is general regional air quality monitoring. The latter is useful for public alerts; the former is not.

It would helpful if there was a publicly available comparison of these competing systems, so that the public could get an idea of what’s involved, and express their concerns to their elected representatives. The LA Blog is only a small effort to fill that need; but hopefully this has helped you to understand what’s going on.