Oral argument — Sierra Club v. EPA

Case numbers 23-3581 and 23-3583. Sierra Club v. United States Environmental Protection Agency et al. Oral argument not to exceed 15 minutes per side. Nicholas Leonard for the appellant, you may proceed. Counsel, if you could please remember to tell us how much you're reserving for rebuttal. Of course, Your Honors. Good afternoon, Your Honors. My name is Nick Leonard on behalf of the Sierra Club, and we would like to reserve three minutes for rebuttal. Keep your voice up, please. Keep my voice up? Yeah. Okay. This case involves two intertwined decisions by the EPA, both of which were unlawful. First, EPA's clean data determination, which relied on its exceptional event determination to exclude two days of high ozone pollution in the Detroit area due to alleged wildfire smoke impacts. And second, EPA's re-designation determination. Your Honors, I will be presenting the argument that EPA's exceptional event demonstration was unlawful because it was arbitrary and capricious. My colleague, Ms. Saxhouse, will be presenting argument regarding EPA's re-designation determination. Before I get to the bulk of my argument, I want to underscore the real-world implications of this case. The time EPA made these decisions, and continuously until today, ozone pollution in Detroit has been above the level that exists to protect people's health. EPA's actions have functioned to strip Detroit residents that live near oil refineries and auto assembly plants of the protections of the Cleaner Act that typically apply when there is an unsafe level of ozone pollution. Here, EPA has acknowledged that Detroit has some of the highest levels of ozone pollution found anywhere in the country. Additionally, the census tract where the monitor at issue in this case is located has the highest asthma hospitalization rate found anywhere in the state of Michigan. To put it bluntly, Your Honors, EPA's decision in this case will result in further exacerbation of an already severe and growing public health problem. I guess maybe I don't understand the process, but if they're out of compliance, then I thought then they'll be declared out of compliance. Yes, Your Honor, that is the normal course of procedure, but the exceptional event demonstration, which we argue was arbitrary and capriciously approved, basically excluded two days of high ozone levels and functionally... That's not what I was asking. I thought you were saying that the levels... When you referred to the levels being the highest and the pollution being... Are you talking about back in those years or now? We're talking about back in those years and now, and so I think it's important to realize sort of essentially the disconnect between essentially what the air quality monitor is measuring and what the sort of design value is after EPA has excluded those two high ozone days. So there's kind of a disconnect that's created obviously if you toss out two days of high ozone levels and thus lower the standard that's used to determine regulatory compliance, but the elevated levels of ozone are still there. Elevated what? Elevated levels of ozone pollution are still in existence, but sort of the legal determination, the legal threshold that you're using has changed via the exceptional event demonstration. Okay. On that point though, let me ask you. So you say that it's arbitrary and capricious. However, it seems that the EPA did use modalities for determining whether or not there was the exceptional event. I'm trying to remember all of the terminology. Was the cause of elevated levels. It seems like the methods that they used were acceptable methods. And so I have a hard time, I get that you're saying, yeah, those methods were acceptable, but look, they didn't give us this and we don't know why these two days were taken out. And that seems more to me like a question about maybe not liking what they did as opposed to saying that what they did really didn't have a basis and was arbitrary. I guess maybe if you could drill down on that for me. Sure, I'd be happy to. And Your Honor, to be clear, we argue that EPA's determination was arbitrary and capricious for three specific reasons. And I'm going to highlight the first of them, but I'll run through all three. EPA has provided conclusions that are not supported by record evidence. It has failed to provide more than conclusory and dismissive statements regarding contradictory evidence. And it has failed to examine relevant evidence. And so what I wanted to focus on today was EPA's consideration of on-the-ground pollution data for fine particulate matter and brown carbon because I think those are the most relevant. And that data is particularly important because generally when you see wildfire smoke present in the area, wildfire smoke is made up of a range of small particles and gases such that where there is wildfire smoke, there are also elevated levels of a wide range of pollutants. And our primary issue here, which I want to present to you, is essentially there were no elevated levels of any of those pollutants of fine particulate matter and brown carbon, which the EPA looked at. And regarding fine particulate matter, EPA even acknowledges that there are no elevated levels of fine particulate matter during the exceptional event days. But didn't EPA, in terms of data collection, find that there were these measurements of the brown carbon and brown carbon, that there were sort of byproducts of incomplete combustion, sort of indicative of wildfire smoke as opposed to pollutants? They did. And our contention is that that finding was simply not supported by record evidence. Because if you look at record evidence, it basically shows that brown carbon was actually highest the day before the exceptional event and then was generally low during the exceptional event. And your honors, the basic premise here is that where you find wildfire smoke, you also find elevated levels of those pollutants. But it sort of seems like you're, you know, there's a debate a little bit over the science of it all. And that you got one position, EPA has another position. But that, I don't know if that speaks to the EPA not acting within a zone of reasonableness. Sure. Our argument is that these aren't scientific determinations. I think what that argument would look like is if we were arguing over the type of pollutants that EPA relied on or the probative value of a particular type of evidence. Your honors, I think it's pretty clear that we are simply arguing under a straight arbitrary and capricious standard that the record, that the evidence in the record simply did not support EPA's conclusion. And that it in fact contradicted them at numerous points. And I see my time is up, but if anybody has any questions, I would simply close by saying that we request that this court vacate EPA's clean data determination and its exceptional event determination. Thank you, your honors. Good afternoon, your honors, and may it please the court. Alina Saxonhouse on behalf of Sierra Club. I'm going to address EPA's redesignation decision and in particular subsection 5 of the redesignation criteria. But to be clear, if the court finds for Sierra Club on the exceptional event issue just discussed, the parties agree the court must vacate the redesignation as well. Under subsection 1, it would not need to reach the other criteria. I do want to answer the question that you asked, Judge White, before I go into my argument. Because I think you're asking why, if the ozone levels are elevated right now, would EPA not just determine that it's a nonattainment. And because they've redesignated it, at that point, even if ozone levels are above the standard, it doesn't automatically revert back to a nonattainment designation and there's nothing requiring EPA to do analysis. Because they've redesignated the area to attainment, at this point, the maintenance plan that the state has put in place governs. And if the ozone levels are above the standard, all that does is trigger the contingency measures in the plan. And that's part of why it's so important that they require these other measures first, which I'll get to. If the court does reach subsection 5, its task under Loper-Bright is to determine the best reading of that provision. Subsection 5 requires EPA to determine that the state has met all requirements applicable to the area and references provisions that include the nonattainment area requirements. EPA agrees that the requirement to adopt pollution control limits, known as RACT, applied to the Detroit area at the time of its redesignation decision. That's on page 60 of its brief. And that Michigan did not comply with the requirement. But EPA argues that the best reading of the statute allows for a loophole based on the timing of the state's application, so that the state is exempt from adopting RACT rules because that requirement came due after its application. But this cannot be the best reading where it is unsupported by subsection 5 itself and conflicts with other provisions that are fundamental to the Clean Air Act 1990 amendments, overall statute, and other provisions. This is a scheme to protect the public from dangerous levels of ozone pollution. The 1990 amendments set out to limit EPA and state's discretion and instead impose bright line requirements for areas that fail to attain ozone standards. This is reflected in sections 7511B2 and 7511A, subsection B. Do you have examples of where it's been interpreted in the way that you are saying that it should be interpreted here? No, EPA has interpreted in their way. They have been consistent, but their interpretation is not rooted in the text. And that's the fundamental problem with it. It's more of a policy preference. The Clean Air Act 1990 amendments make it automatic that when a marginal non-attainment area fails to attain by its deadline, EPA must reclassify it to moderate. And automatic that when an area is reclassified to moderate, stricter requirements like RACT kick in. In other words, they explicitly remove the regulator's discretion to decide whether additional measures are needed for areas that are struggling to maintain healthy air quality. A reading that allows EPA to unilaterally reassert that discretion doesn't make sense. The mandatory shall language in the RACT requirement in section 7511A was the key reason that this court in both Wall v. EPA and the 2015 Sierra Club v. EPA case found even applying Chevron deference, EPA could not create loopholes for the RACT requirement. One other key point of statutory context is the relationship between subsection 5 and the ongoing protection of public health, even once an area is redesignated. The RACT rules limit pollution from things like vehicle coating operations, refineries, asphalt mixers, and they're not busy work. They have a continuing role after redesignation if they're adopted prior. They become part of the maintenance plan for the area. And even if the state seeks to remove them as part of that baseline plan, at a minimum, they must remain as a contingency measure if ozone levels start violating the standard again. That's required by 7505AD, and this point was recognized by the court in Wall. And those contingency measures are key. This is what I was talking about before, because once redesignated, air quality violations don't revert the area back to the stricter nonattainment regime. It's all up to the contingency measures at that point to get the area back to attainment. So when EPA doesn't enforce the RACT requirement as a result of its subsection 5 reading, the maintenance plan that carries forward is weaker and less protective of public health. So this is a real-world consequence for Detroit. I want to, in my last minute, address some of EPA's counterarguments. So EPA claims it would miss its 18-month deadline to approve or deny the RACT requirement. It would also deny the redesignation request if it had to wait for a state to complete new requirements after a bump-up. And the bump-up is the reclassification to moderate. That's sort of the shorthand. But bump-ups are not a surprise. States can and do prepare ahead of time to meet their requirements. For example, Michigan began working on its RACT rules five or six months before they were formally required. EPA also can and does at times act more quickly to review state submissions than it acknowledges. And we've provided examples of that on page 46 of our reply. EPA could also meet its deadline by denying the state's requests until it can complete the new requirements or put it on hold. And if the area is attaining, still meeting the standard, it can grant a clean data finding to give the state relief from other non-attainment requirements in the meantime. And most importantly, even if there were some tension with the deadline, that would not dictate that EPA can jettison a different mandatory statutory requirement there. Thank you. Heather Gangy My name is Heather Gangy. I'm with the U.S. Department of Justice. And I represent EPA in this case. And with me at council table is Elizabeth Pettit, EPA counsel on this case as well. As my colleagues indicated, this case concerns two overlapping EPA rulemakings for the Detroit area. A clean data determination, finding that the Detroit area's air quality was in compliance with the 2015 ozone NAAQS through the time that rulemaking was finalized. And also the redesignation of the Detroit area to attainment of that air quality standard. These rules were reviewed under the arbitrary and capricious standard. And under that standard, they should be upheld because EPA did consider relevant data. There is a robust record basis for the determinations that it made. And EPA explained the relationships between the data and the analyses it considered and the conclusions that it reached. Many, if not most, of those determinations also are complex scientific determinations. And they are entitled to substantial deference. Let me ask you one question about that. So when we talk about the science, it seems to be competing views on how to view the data from a scientific perspective. And in terms of the two days of exceedance, the June 24, 25, I think it was, doesn't there have to be like a clear causal relationship between the exceedance and the wildfire smoke? It seems to hear that, I guess I wonder if it's clear. It just seems to be, maybe it's just debatable. It is EPA's position that there is a, first, yes, a clear causal relationship is required. That is required by the statute. Ms. Gange, I know that probably to you, you sound like you're very loud. We're having difficulty hearing you. If you could just speak a little bit louder. Okay. Sorry about that. Can you hear me now? It's better. Thank you. I apologize. Yes, both the statute and the implementing regulations require a clear causal connection between the extraordinary event, which was Canadian wildfires, and exceedance of the ozone standard in Detroit, which was the exceptional event. Do we have, what was the data on, you've got the days when, you know, the plaintiffs assert the wildfires were the worst. Apparently, the numbers were not elevated. Is that correct? And then the next two days, the numbers are elevated. Is that correct? That is, I think, what Judge Cole has been pointing out as a point of scientific disagreement. Well, but I have a question, a factual question. Does the record show what the levels were the following two days? The record, what the record shows, and what EPA explained very fulsomely, and so did Michigan in its demonstration, is that wildfire smoke from Canada was drawn down on June 23rd, which is, as the petitioners point out, was the day that brown carbon levels were the highest in Michigan. Because that was the day that a plume of wildfire smoke was drawn into the area. It was drawn in at high elevation. On the 24th and 25th, a lot of that smoke actually fell down to the ground, because of the meteorology on that day. No, I understand that. And so ozone levels were not as high on the 23rd, when most of the smoke was up at high levels. But on the 24th and 25th, when that smoke fell down to the ground and reached the monitors, because the air monitors are on the ground, that's when the smoke, mixing with other emissions in the area and the heat of the day. But I'm asking, what were the readings after those two days? I am not aware of that. But after the 25th, the ozone levels were not high enough to cause the area to not attain. The ozone levels were highest on the 24th and 25th, and those were so high that those days, if they had not been taken out under the exceptional event determination, they would have caused the Detroit area not to attain the ozone standard for 2022. And are there other, if one were to look at the records for the year, are there other times when there were these isolated peaks, which is how I think you're describing this? No, there were not other days that were affected by wildfire smoke. I don't mean wildfire. I mean other, if you're just looking at the readings and not extrapolating a cause, but if you just look at the readings, do the readings look like, you know, if you were to graph it, does it look like that, or is it basically even and then there was a big spike? There are no other days in 2022 where the ozone levels were so high that they essentially spiked up to such a high level that they would have caused Detroit not to attain the 2015 standard. It's not really an answer because you could have 10 spikes and it's the 11th that puts you over. The way attainment is measured for these air quality standards, you look at data that's collected by air monitoring stations across the area. I believe in Detroit there are eight or nine different monitors. You look at each monitor separately and you take an average of the fourth highest reading and you use that average to determine whether or not air quality is exceeding. So saying that these two days put you over doesn't answer the question whether, if you look at this monitor for a period of, say, nine months, do you get readings that are going up and down or do you get a pretty basic level and then on these days they shot up? Across from day to day, ozone levels do go up and down at every monitor because to get ozone, sources don't just emit ozone. It doesn't come out of power plants or car tailpipes. You have to have other chemicals they call precursors like oxides of nitrogen and volatile organic carbons. They have to mix together, the weather has to be warm enough, there has to be enough heat and there also needs to be a certain level of humidity in the air. And when all of those factors come together, the chemicals interact to create ozone and the monitors register ozone. So how much ozone there is can vary by humidity, it can vary by temperature and it also can vary by how many of those precursor chemicals you have in the area. So levels do go up and down because all of those variables change from day to day but here you have a monitor that was measuring ozone levels at or below the standard except for two days in June where wildfire smoke brought in ozone precursors in addition to what had been present in the Detroit area and between those extra precursors, the heat of the day in June and humidity level, those chemical reactions occurred and there was enough ozone measured to put the area over the standard. So there were spikes on those days that were high enough to take Detroit out of attainment, there likely have been spikes on other days too but not so high as to actually violate the air quality. And what about the argument that the chemicals that usually accompany wildfires were absent? That is not an accurate statement for two reasons. I would direct the court to page 127 of the joint appendix where Michigan explained this pretty clearly in its submittal. There are different monitors in the Detroit area measure different things and other monitors in the Detroit area were monitoring different components of wildfire smoke. So are we only talking about though, and I might have missed some of that, there's a lot of talk about the E7 mile monitor and then I think there was a reference to what was being registered in Dearborn, are there other monitors around the area that were picking up on some of these other things that were indicative? Yes, yes there were. That is discussed at page 127 of the joint appendix. Pages 122 and 123 of the joint appendix also show plume maps that were generated by the National Oceanic and Atmospheric Administration or NOAA among many other things that EPA considered. Those maps show very large plumes of wildfire smoke coming down into the area from Canada. And then the different areas... So my understanding is that the plume is not actually, the plume smoke is kind of a visual determination and it's not something that is, and you correct me if I'm wrong because I'm no scientist here, it's not measuring actual chemicals, is that right? The plume itself, like you would see on satellite or other maps like at pages 122 and 123 of the joint appendix, the smoke itself that you see is a mixture of a lot of different chemicals and substances. The air monitors on the ground measure particular things. They're manufactured, like some will measure particulate matter, others might measure volatile organic carbons or ozone. So the monitors are not looking for smoke, they're just looking for particular chemicals. And here there is quite a robust record showing that there was smoke that came down from Canada and affected the area. There's robust evidence in the record that that smoke actually did fall to the ground and affect air monitors, including the seven-mile monitor that exceeded the NAAQS. And then there is additional evidence, including a matching day analysis and a meteorological analysis that EPA did. Pulling all of the evidence together to show that the wildfire smoke actually did cause ozone levels to spike on the 24th and 25th. But the Sierra Club disputes the points you have just made. I mean, there are two sides to this coin. And Sierra Club looks at the same information that EPA looked at, and they would like EPA to draw a different conclusion. But that is not how rulemaking should be reviewed under the arbitrary and capricious standard. A rule can only be overturned or set aside by the court if the agency acted arbitrarily and capriciously. And the agency does not do that where it clearly identifies the information that it considered and explains the connection between the information that it considered and the conclusions that it reached. And as we've been discussing, where those conclusions are scientific in nature, the agency is entitled to particular deference, especially where that area of science is within its expertise and its regulatory role. And that is precisely the case here. And if the court... Thank you. I would like to move on to the redesignation to attainment. Opposing counsel has challenged three of the five redesignation criteria. EPA does agree that the first of those three criteria stands or falls on the clean data determination, which we have just been discussing. And that is a clear and straightforward analysis, the agency thinks. And so that should be upheld. The fifth, EPA's determination that the fifth criterion was satisfied also should be upheld. The parties do not dispute that racked or reasonably available control technology is not required for marginal attainment areas. And the Detroit area was a marginal non-attainment area at the time that Michigan submitted its request to be re-designated to attainment. So it looks like your time is up, if you want to wrap up. Yes, Your Honor. I think in summary for that last criterion, I believe that the agency laid out in its briefing why it believes that its long-standing interpretation is the most appropriate. I have one question. If I understand the facts correctly, I'm just a little confused by the dates. A deadline is set. In June of 2018, they're given the deadline of August 3rd, 2021 to reach attainment, right? I believe so, Your Honor. They submit an application in January of 2022. And in March, it's proposed to designate that they reached attainment. But I don't understand how it's decided that they didn't meet the 8-21 deadline if the EPA is agreeing that they're in attainment. I'm just not following how they could be not in attainment and then all of a sudden they're in attainment. It is confusing with the time frames. And there were several different things going on at the same time. The bottom line is that in order to attain one of the NACs, or the National Ambient Air Quality Standards, the area's air quality actually has to be comparable. And it has to be compliant for three years in a row. For how many? Three. Three years in a row. The Detroit area's air quality did not comply with the standard in 2018, and its deadline was 2021. So to timely comply with the standard, it had to have been... considered attaining by the deadline. Its air quality was not good enough in 2018. It was good enough in 2019, 2020, 2021, and 2022. But because it had only complied with the standard for two years before 2021, EPA was forced to find that it did not timely attain. Which would seem not to make sense, right? Because its air quality actually was good enough in 2021, but the agency had to say, no, you didn't meet the standard soon enough. So that's why they were found not to have timely attained at the same time that EPA was entertaining a request to redesignate them to attainment of the standard in 2022. Which they could do, because the air quality was good enough in 2019, 2020, and 2021. Does this happen a lot? It's not an unusual situation for areas to... Their air quality is good enough, but it's not good enough soon enough. And so they get stuck in this exact kind of situation that the Detroit area is stuck in. And for three years, you sort of kind of are relieved from those earlier obligations. The way the Clean Air Act works is you only become relieved of a number of obligations after an area is formally redesignated to attainment. So to be formally redesignated, the five redesignation criteria that are discussed in the brief all have to be met. There is a very substantial, complicated, substantive package that has to be given to EPA. And EPA has to examine and then approve it. That takes a really long time. It can take so long to get redesignated that areas can find themselves continually getting bumped up, even when their air quality is perfectly fine. So there's a second option, which is also at issue in this place, in this case called the clean data determination, where a state can go... It sounds like this might be going beyond the question. Do you want to hear that? I guess... So the procedure that happened here was the clean data determination, right? Yes. So my question is, does it happen often that an area becomes compliant and they take that route and you have these seemingly contradictory determinations? It is not unusual for states to ask for a clean data determination, because it relieves them of a few obligations. Okay. All right. Just a few. Nowhere near as much as they obtain when they're formally redesignated. Thank you. Thank you very much. I have a quick logistical question. Would it be permissible to split our rebuttal time between the two councils? What would be the split? The three minute rebuttal time that we preserve. One and a half, one and a half. I can't do one and a half. I think we need whole numbers. Two it is. I just want to quickly underscore a couple of points here. First, council for EPA stated that the ozone levels in Detroit haven't exceeded the standards on other days. That's simply not true. It has, in fact, exceeded the standards on other days. That's a really important point. Detroit is obviously a highly industrialized area. There are a number of local pollution sources that contribute to high ozone levels, which is why we've seen that continuing problem in the past and through present day. And so the question here is basically, okay, well, what caused the high ozone levels on June 24th and 25th? Was it local pollution sources? Was it this wildfire smoke? To be clear, our argument is that the evidence points to wildfire smoke not being present in the area on those two days. I want to bring your attention to one of the points EPA's council made, which is that on June 23rd, there was no smoke at ground level. That smoke was still in the upper atmosphere. But, your honors, and this point is not in contention, brown carbon data measured at Dearborn was at its highest on June 23rd when EPA claimed smoke wasn't even present at the ground level. And then it was at its lowest on June 24th and 25th when EPA claims that smoke descended to the ground level and affected the air quality that people breathe. That's not a scientific dispute. That's just, simply put, EPA making a statement that's not connected to the factual record. That point is underscored by the lack of fine particulate matter data, or PM2.5, throughout the exceptional event period, which, once again, EPA doesn't dispute. There are no elevated levels of fine particulate matter, which is a common indicator of wildfire smoke. And EPA acknowledged that, Michigan acknowledged it, and it is, once again, evidence that wildfire smoke simply wasn't present in the area. Just a real quick question, and I know your time is tight. Michigan used this Lake Michigan Air Directors Consortium screening analysis. Yes, that's the PM2.5. Indicated that there was smoke present. Your Honor, that actually, and Michigan acknowledged it, did not indicate that wildfire smoke was present. There's an email that we cite in our brief where essentially Michigan says, this points to wildfire smoke not being present in the area. And if I could close by just pointing to one thing, there's also satellite imagery from June 24th, and this is provided in figure two on page 14 of our reply brief. It shows no smoke in the entirety of the southeast Michigan area on June 24th. These aren't technical points. These are basic disconnects between EPA's conclusions and the facts in the record. Thank you. I want to just quickly respond to EPA's contention that if an area is not redesignated in time, it will continually get bumped up. If ozone levels remain below the standard, the area would not get bumped up. No new requirements would become due. That's because it would only get bumped up if it fails to attain, so if the levels of ozone go back up. So this idea that they can have clean air and keep getting bumped up is not correct. Secondly, I wanted to respond to this idea of the consistency of EPA's interpretation. It's a longstanding interpretation that that would be something that could overcome the text. It can't. Even when longstanding, Loper-Bright and Skidmore instruct that the court should consider the thoroughness of an agency's explanation of the agency's analysis of the text and validity of its reasoning before giving it heightened respect. The Calcagney memo relied upon EPA does not include any textual analysis at all. And even under Skidmore deference, agency interpretations only have power to persuade and not to control. Thank you. The case will be submitted.