A couple of weeks ago I showed how the Natural Resources Defense Council (NRDC) worked with a group of university and government scientists to mislead Americans into thinking that ozone smog levels will rise in the future. The report, Heat Advisory: How Global Warming Causes More Bad Air Days, is an update of a report NRDC first released in 2004. Dan Lashof, the Science Director of NRDC’s Climate Center, responded to my critique last Friday. Here are my comments on Lashof’s response. You can scroll to the bottom to see Lashof’s response in its entirety.
The central point of my original critique is that NRDC and its team of outside climate and health scientists misled reporters and the public by claiming that ozone levels will be higher in the 2050s than they are today and that climate change will be the cause of this increase in ozone. I showed that in reality NRDC generated an appearance of increasing ozone by assuming higher emissions of ozone-forming pollutants in the future. Climate change is a minor factor compared to assumptions about ozone-forming emissions, but NRDC and its scientists obscured this fact. Dan Lashof doesn’t dispute this central conclusion, but he does continue NRDC’s history of evasive and misleading claims about Heat Advisory’s assumptions and about future ozone-forming emissions.
Rather than acknowledge that NRDC vastly exaggerated future ozone-forming emissions, Lashof equivocates: “While we would expect significant reductions in precursor emissions over the next decade there are no reliable estimates of precursor emissions extending to the mid 21st century. Thus different research projects use different sets of assumptions to project future precursor emissions. The project on which Heat Advisory is based kept anthropogenic ozone precursor emission levels constant as a way of evaluating the effect that climate change alone could have on ozone concentrations.”
Let me unpack the false and misleading claims and insinuations here.
First, Lashof says NRDC “kept anthropogenic ozone precursor emission levels constant.” His phrasing could lead readers to the mistaken impression that NRDC held future ozone-forming emissions constant at current levels. In fact, NRDC held emissions constant at the 1996 level, which, as already noted, is more than 30% higher than current emissions and at least four or five times greater than emissions will be in the 2050s under current regulatory requirements (see below for more on future ozone-forming emissions).
Second, Lashof says “While we would expect significant reductions in precursor emissions over the next decade there are no reliable estimates of precursor emissions extending to the mid 21st century.” In other words, NRDC realizes that ozone-forming emissions will decline. But since they don’t know how large the declines will be, they’ll just assume emissions will increase by more than 30%.
Lashof’s special pleading here is particularly ironic. Climate activists have no problem trying to force the people of the world to spend trillions of dollars for CO2 reductions based on the presumption that climate models are accurate. But when it comes to ozone, NRDC pleads uncertainty and then chooses increases in future ozone-forming emissions that are grossly at odds with any plausible future scenario. If anything, the statement that “there are no reliable estimates….extending to the mid 21st Century” is far more applicable to greenhouse gas emissions and climate models’ predictive skill than it is for ozone-forming emissions.
Third, Lashof says “The project on which Heat Advisory is based kept anthropogenic ozone precursor emission levels constant as a way of evaluating the effect that climate change alone could have on ozone concentrations.” Just imagine if climate skeptics used 1996 greenhouse gas emissions levels to project future climate change. NRDC would be all over them like a cheap suit, and rightly so. Lashof seems unaware that using 1996 ozone-forming emissions to predict future ozone levels is equally ridiculous. Just as annual CO2 emissions will continue to rise, ozone-forming emissions will continue to fall.
NRDC got its analysis backwards. They should have started with an estimate of likely ozone-forming emissions in the 2050s, and then estimated future ozone levels with and without warming. That would answer the policy-relevant question. The answer: ozone declines a lot either way, but it declines slightly less in some areas if the climate warms. You can see this by looking at Tagaris et al. (2007), which came closer to using realistic future levels of ozone-forming emissions (though even they overestimate future emissions, as I show here). One might conclude that NRDC and its allied scientists didn’t do the study that way because it wouldn’t have served their political goal of creating an impression that air pollution will increase in the future. Headlines like “Ozone exceedance days will decline 90% instead of 95% if climate warms” don’t generate the kind of fear and outrage NRDC thrives on.
Fourth, just as important as what NRDC did do in its study is what it did not do. Notice that Heat Advisory focuses on ozone levels in the northeastern and southeastern U.S. It turns out that the northeast may be the region where warming would have the largest effect on ozone levels. The story is quite different if you look at the entire U.S. According to Tagaris et al.’s modeling (Table 1a, unfortunately not available online), all else equal, warming lowers ozone levels in the Midwest by nearly as much as it raises ozone in the northeast. For the U.S. as a whole, the net effect of warming alone on ozone levels is zero. In other words, whether intentionally or not, NRDC focused on the region of the U.S. most congenial to the scary story it wanted to tell about warming and ozone levels. Equally important, Tagaris et al. show that while climate is likely to have a small effect on ozone levels in some areas, emissions of ozone-forming pollutants are far more important and swamp any effects of climate change.
What about future ozone-forming emissions? In my original critique, I cited a number of NRDC press releases in which NRDC highlighted the proposal, adoption or implementation, of EPA regulations that will eliminate most remaining ozone-forming during the next couple of decades. Here is Lashof’s response: “NRDC has applauded some recent regulations to reduce precursor emissions and has sued EPA over other proposed regulatory changes that would increase emissions. While we would expect significant reductions in precursor emissions over the next decade there are no reliable estimates of precursor emissions extending to the mid 21st century.”
Lashof could have strengthened his case by providing some specifics about these “proposed regulatory changes that would increase emissions.” He would also need to show that these supposed emissions increases would be sufficient not only to offset the emissions-reducing regulations, but to also cause more than a 30% increase in ozone-forming emissions above current levels. I submit that the reason Lashof does not give specifics is that if he did he would have to admit that ozone-forming emissions are guaranteed to drop in the future, even if we interpret EPA regulations in way that’s as favorable as possible to NRDC’s desires. But if he made that admission, he’d also have to concede that NRDC’s study is based on fraudulent premises. Instead, Lashof chose evasive, equivocal statements that make it appear as if it is difficult to discern the net effect of EPA requirements on future ozone-forming emissions. Actually, it’s quite easy.
Most ozone-forming emissions come from motor vehicles and power plants. As I showed in my original critique, NRDC has put out press releases highlighting EPA regulations that require at least a 90% reduction in per-mile emissions of NOx and soot from on- and off-road diesel vehicles; a 77%-95% reduction in NOx and VOC emissions from automobiles (including SUVs and pickups, and even Hummers, with the larger percentage reductions applying to the larger vehicles); and a 50% reduction in NOx from coal-fired power plants (actually EPA claims the reduction is more than 60% below 2003 levels).
But there’s more. Under Title III of the Clean Air Act EPA has imposed dozens of Maximum Achievable Control Technology (MACT) requirements for dozens of industrial sources. The MACT requirements eliminate most emissions from dozens of different categories of industrial sources, such as refineries and organic chemical manufacturers. And there are still more rules for various and sundry other sources of air pollution, such as reductions in emissions from paints and coatings and consumer products, various types of specials engines and vehicles, locomotives, commercial establishments, and on and on. Furthermore, many states, including California, North Carolina, and several northeastern states, have adopted emissions requirements that are even more stringent than EPA’s. And as NRDC knows full well, many more regulations will be coming down the pike over the next decade—indeed, I’d bet that NRDC is lobbying for such regulations even as I write this—as America’s metropolitan areas struggle to comply with EPA’s ever-tougher ambient standards for ozone and other air pollutants.
We also have tons of direct evidence that emissions are steadily dropping. For example, under EPA’s NOx SIP Call regulation, coal-fired power plants reduced May-September NOx emissions (that’s the time of year NOx matters for ozone formation) by 57% between 1998 and 2006. On-road measurements from the mid-1990s through 2004 show that VOC emissions of the average automobile have been dropping about 12% per year, CO by about 10%/year, and NOx by about 6%/year (I show the evidence through 2001 here; I can provide more recent evidence on request). And that’s before the 2004 implementation of the “Tier 2” automobile emissions standards which, by NRDC’s own admission, will reduce emissions of 2004-and-newer automobiles by 77-95% below the emissions of previous models. A new study from scientists at UC Berkeley (in press in the journal Atmospheric Environment, but not yet posted online) finds that diesel truck NOx emissions per mile declined about 30% between 1997 and 2004. These improvements will speed up as the 2007 emissions standards come into force.
We can also see real-world reductions in ozone-forming emissions by looking at ambient air measurements, which show that NOx, CO, total VOC, and specific VOCs like benzene and 1,3-butadiene continue to decline by a few percent to several percent per year around the nation.
Since existing federal regulations already require the elimination of most remaining ozone-forming emissions, where are these “proposed regulatory changes that would increase emissions” that Dan Lashof refers to but about which he provides no specifics? The only thing he could mean is changes to New Source Review (NSR)? Environmentalists have spent the last several years in a successful campaign to convince reporters and the public that reforming NSR would increase industrial emissions. But this claim was and is false, because caps on total emissions from power plants and industry under Title IV and Title III of the Clean Air Act, as well as the new, even tougher caps under the NOx SIP Call and the Clean Air Interstate Rule, remain in effect regardless of any changes to NSR. If there are any policies that would actually increase emissions, I would ask Dan Lashof and NRDC to please reveal specifically what they are, and also to explain how any such policies would offset even a tiny fraction of emissions reductions that are coming down the pike from existing regulations.
Overall, suffice it to say that existing regulations and requirements will eliminate at least 70% to 80% of remaining ozone-forming emissions during the next two decades or so. Yet NRDC assumed more than a 30% increase in ozone-forming emissions between now and the 2050s. Once again, NRDC predicts higher ozone in the future, not because of climate warming, but by assuming ozone-forming emissions will be at least four or five times higher than they actually will be.
In defending the assumptions NRDC chose, Lashof says “Other researchers may choose alternative assumptions about how anthropogenic ozone precursors could change in the future, and will arrive at different projected ozone concentrations.” Yes, indeed, you can make any assumptions you want about future ozone-precursor emissions. But Lashof omits the fact that NRDC chose assumptions that are patently at odds with reality. If realistic assumptions would have generated increases in future ozone, NRDC and its allied scientists would probably have done a realistic study. But the study wasn’t done as an inquiry into likely future ozone levels. Rather, the outcome—higher future ozone—was chosen up front. Then the assumptions were chosen so as to generate the desired outcome.
Despite the media blitz surrounding the release of Heat Advisory
, Lashof now claims that NRDC wasn’t making any predictions at all about future ozone levels. He says “Projections of how global warming would affect ozone levels are not predictions of what will happen.” That would be news to anyone who read NRDC’s press release
announcing Heat Advisory
. Here’s how it begins:
New Study: Smog Poses Greater Health Risk Because of Global WarmingMore Bad Air Days for Southern, Eastern U.S. Cities
WASHINGTON, DC (September 13, 2007) — People living in ten mid-sized metropolitan areas are expected to experience significantly more ‘red alert’ air pollution days in coming years due to increasing lung-damaging smog caused by higher temperatures from global warming.
No predictions there.
How about these lines from later in the press release:
Researchers project that, unless action is taken to curb global warming, by mid-century people living in a total of 50 cities in the eastern United States would see:
- A doubling of the number of unhealthy ‘red alert’ days
- A 68 percent (5.5 day) increase in the average number of days exceeding the current 8-hour ozone standard established by the U.S. Environmental Protection Agency (EPA)
- A 15 percent drop in the number of summer days with “good” air quality based on U.S. Environmental Protection Agency (EPA) criteria because of global warming
I guess those aren’t predictions either. Oh, right, they’re “projections,” not “predictions.” Now there’s a distinction without a difference. Either way, NRDC has misled people into the mistaken belief that ozone will increase in the future. Yet now that the damage is done and Americans have been snookered, Lashof is trying to weasel out of taking responsibility for NRDC’s unfounded scare-mongering.
Heat Advisory is a case study in how scientists can generate a pre-determined result—in this case, the false appearance of rising ozone levels—by careful management of assumptions and analytical boundaries, by crafting statements that are true on their own but that lead readers to draw false conclusions, and by shrewd choices about what information to omit from a paper.
Here are a few ways that these efforts played out in Heat Advisory:
- Assuming higher ozone-forming emissions in the future guaranteed that Heat Advisory would produce rising ozone levels, regardless of the effect of climate change.
- NRDC says it “kept ozone precursor emissions constant”—a true-but-misleading statement, because “constant” means constant at the level of 1996, which is more than 30% higher than current emissions. Furthermore, there’s no mention of the reality that ozone-forming emissions will in fact sharply decline.
- NRDC limits its analysis to areas of the country where the models say that, all else equal, warming will increase ozone, but omits areas where the models say warming reduces or has no effect on ozone.
- NRDC also omits the fact that the effect of climate warming on ozone levels is swamped by the effect of ozone-precursor emissions on ozone levels.
While all of the above examples are illuminating, the most profound flaw in Heat Advisory is that its misleading results were built in right from the outset in the very question NRDC chose to ask, namely: Given “current” ozone-precursor emissions (really 1996 emissions), how will climate warming affect ozone levels in the 2050s? In effect, this really amounts to asking how much higher ozone would have been back in 1996 if the temperature had been a few degrees warmer—a counterfactual scenario of no relevance for climate or air pollution policy. The beauty of NRDC’s sleight of hand is that Heat Advisory appears to be asking a question about future ozone levels when in fact it is asking a question about past ozone levels.
The real policy-relevant question is this: Give likely ozone-precursor emissions in the 2050s, how will climate warming affect ozone levels in the 2050s? A question neither asked nor answered by NRDC.
NRDC and other environmental groups style themselves as arbiters of scientific probity in environmental debates. Examples such as Heat Advisory show that these groups and their university and government allies can be as crassly political in their use of scientific authority as they imagine their opponents to be. (Here is an example of similar behavior by the Union of Concerned Scientists.)
Heat Advisory is a cynical exercise in manipulation and fear-mongering masquerading as science. Perhaps we shouldn’t be surprised that NRDC manufactures bogus increases in ozone in order to scare people (and then dissembles further when someone like me calls them on it). After all, environmentalists derive their power from public fear and outrage. Where there is no legitimate reason for fear and outrage, environmental groups and some scientists have shown themselves to be only too willing to manufacture it.
APPENDIX I: Additional ways in which Heat Advisory misleads Americans
I haven’t yet discussed NRDC’s health effects claims in my Heat Advisory critique. Of course, the purpose of Heat Advisory is not merely to create the misimpression that ozone will rise in the future—the ultimate goal is to foment greater fear of harm from air pollution and to link that fear to climate change. But just as NRDC exaggerates ozone levels, it also exaggerates the harm from any given level of ozone. I give details, with citations to the literature, here, here, and here.
I also haven’t yet discussed the predictions of climate warming on which Heat Advisory’s ozone predictions are based. In fact, NRDC used a greenhouse gas emissions scenario that likely exaggerates future warming. This is true even if we assume that climate models do a good job at predicting future climate. The researchers who wrote NRDC’s report and the companion journal article in Climatic Change used the IPCC’s A2 scenario of future greenhouse gas emissions to predict future climate change. A2 is the second warmest of the IPCC’s six scenario groups, but it is also one of the least plausible. Here’s why: A2 assumes that world population continues to rise steadily, reaching 15 billion by 2100. But the International Institute for Applied Systems Analysis (IIASA) projects, based on world fertility and life expectancy trends, that world population will peak just above 9 billion around 2070 and decline thereafter. In fact, the A2 scenario’s population assumption for 2007 already overestimates actual world population by about 275 million people, or more than 4 percent. The discrepancy only grows larger with time as the A2 scenario continues to add greenhouse gas emissions from those non-existent 275 million people and the children and grandchildren those non-existent people will never have.
The A2 scenario includes other assumptions that exaggerate future greenhouse gas levels. For example, A2 assumes atmospheric methane levels steadily rise after 2000 and nearly double by 2100. But actual atmospheric methane concentrations have been steady since 1999. The A2 scenario’s assumption for 2006 is already nearly 6% higher than the measured level. A recent study of methane emissions concluded “Seeing that the total source has remained constant for at least the last two decades, it is questionable whether human activities can cause methane concentrations to increase greatly in the future.”
After carbon dioxide and methane, tropospheric ozone is the next most important human-caused greenhouse gas. But methane helps form ozone, so overstating methane levels also causes an overestimate of ozone. Likewise, NRDC’s assumption of rising ozone-forming emissions exaggerates future ozone levels, thereby overstating greenhouse warming.
APPENDIX II: NRDC’s Full Response to My Original Critique
Here is our response to Joel Schwartz’s attack.Please feel free to forward it to anyone who has received Schwartz’s email.-Dan
Our report is based on a peer-reviewed paper by Bell et al. (2007) published in the scientific journal Climatic Change. See http://www.nrdc.org/globalWarming/heatadvisory/contents.asp
The Bell et al. (2007) research is based on the New York Climate and Health Project, one of the largest-scale ozone modeling efforts conducted in the US. This research applied computer model simulations to project ozone conditions for the entire eastern US over a five-year period in the 1990s (1993-1997) to describe “current” ozone conditions, then compared those concentrations to future projections for a similar five year period in the mid-21st century (2053-2057) under a changing climate.
The analysis of climate impacts on ozone kept anthropogenic ozone precursor emission levels constant as a way of evaluating the climate change effect on ozone concentrations. Because of its complex formation chemistry, ozone concentrations don’t necessarily change in direct proportion to precursor changes. In spite of some recent reductions in precursor emissions, nationwide ozone levels have decreased only slightly (9%) since 1990. (Source: US EPA, Ozone Air Trends, available online at http://www.epa.gov/airtrends/ozone.html)
NRDC has applauded some recent regulations to reduce precursor emissions and has sued EPA over other proposed regulatory changes that would increase emissions. While we would expect significant reductions in precursor emissions over the next decade there are no reliable estimates of precursor emissions extending to the mid 21st century. Thus different research projects use different sets of assumptions to project future precursor emissions. The project on which Heat Advisory is based kept anthropogenic ozone precursor emission levels constant as a way of evaluating the effect that climate change alone could have on ozone concentrations. Other researchers may choose alternative assumptions about how anthropogenic ozone precursors could change in the future, and will arrive at different projected ozone concentrations.
Projections of how global warming would affect ozone levels are not predictions of what will happen. Our goal and our hope is that both global warming pollution and ozone precursor pollution will be reduced fast enough and far enough to avoid the risks we describe.
Daniel A. Lashof, Ph.D.Science Director, Climate CenterNatural Resources Defense Council1200 New York Ave., N.W.Suite 400Washington, DC 20005Phone: 202 289-6868Fax: 202 789-0859Email: firstname.lastname@example.org___________________________________________