In 1997, for instance, the E.P.A. moved to reduce surface ozone, a form of pollution caused, in part, by emissions from oil refineries. Susan Dudley, an economist who became a top official at the Mercatus Center, criticized the proposed rule. The E.P.A., she argued, had not taken into account that smog-free skies would result in more cases of skin cancer. She projected that if pollution were controlled it would cause up to eleven thousand additional cases of skin cancer each year.
In 1999, the District of Columbia Circuit Court took up Dudley’s smog argument. Evaluating the E.P.A. rule, the court found that the E.P.A. had “explicitly disregarded” the “possible health benefits of ozone.” In another part of the opinion, the court ruled, 2-1, that the E.P.A. had overstepped its authority in calibrating standards for ozone emissions. As the Constitutional Accountability Center, a think tank, revealed, the judges in the majority had previously attended legal junkets, on a Montana ranch, that were arranged by the Foundation for Research on Economics and the Environment—a group funded by Koch family foundations. The judges have claimed that the ruling was unaffected by their attendance.
Read more http://www.newyorker.com/reporting/2010/08/30/100830fa_fact_mayer?currentPage=7#ixzz0xT7UXCY9
Natural processes contribute in the springtime to O3 concentrations measured at the surface in the United States at both high and low-elevation monitoring sites and hourly average concentration ≥ 0.05 ppm in the springtime can be attributed at times to these natural processes (Lefohn et al., 2001; Cooper et al., 2005).
10. The value of the W126 and SUM06 cumulative exposure indices is correlated with the large number of hourly average concentrations ≥0.10 ppm (N100) that caused the
2vegetation growth reductions observed in the NCLAN-type experiments. However, under ambient conditions, large numbers of hourly average concentrations ≥0.10 ppm (N100) are not experienced at most monitoring sites across the United States and the value of the ambient cumulative exposure index (e.g., W126) is not correlated with elevated hourly average concentrations. Because of this observation, neither the W126 nor the SUM06 metrics will provide consistent predictions under ambient conditions (i.e., higher values for the metrics do not necessarily imply a greater effect) and will not produce reliable predictions of vegetation effects without including the N100 index.
In other words, the exposure- response relationships are tightly linked to the artificially high number of elevated O3 concentrations that originated in the NCLAN experiments. The elevated hourly average concentrations observed in the NCLAN experiments were responsible for causing the reduction in observed yields. In turn, the elevated hourly average concentrations were highly correlated with the value of the cumulative exposure index (i.e., W126 or SUM06) calculated. Under ambient conditions, large numbers of hourly average concentrations ≥0.10 ppm are not experienced at most monitoring sites across the United States and the value of the cumulative exposure index (e.g., W126) is no longer correlated with the elevated hourly average concentrations that caused the growth reductions observed in the NCLAN-type experiments. Because of this observation, both the W126 and SUM06 metrics are not consistent (i.e., higher values for the metrics do not necessarily imply a greater effect) under ambient conditions and will not produce reliable predictions of vegetation effects without including the N100 index. Yun and Laurence (1999) showed that the same SUM06 value resulted in very different foliar injury when exposure regimes with different numbers of high concentrations were applied. The authors concluded that the SUM06 might not be a good predictor of injury. Additional concerns have been raised in the literature (e.g., Lefohn and Foley, 1992; Lefohn et al., 1997; Edwards et al.,
211994; FLAG, 2000; Musselman et al., 2006) about the use of only the W126 or SUM06 exposure-response relationships for predicting vegetation effects if the relationships were derived using NCLAN-type experimental data. Lefohn (1998) and Lefohn et al. (1997) have provided examples of the application of a multi-component index (i.e., W126 and N100) using data derived from NCLAN fumigation protocols. The authors believed that the multi-component index addressed the problem of the inconsistency of the W126 and SUM06 cumulative exposure indices.
Recently, Davis and Orendovici (2006) confirmed that under actual ambient field exposure conditions, the use of the multi-component exposure index was required. Testing various models, the authors concluded that the model that used (1) plant species, (2) Palmer Drought Severity Index, and (3) the interaction of the W126 exposure index and the N100 index was superior in performance for establishing a statistically significant relationship between vegetation symptoms and O3 exposure than models that did not use the combined W126 and N100 metrics. The authors reported that the incidence of O3 symptoms was most related to the O3 metrics W126 index coupled to the N100 index.
Although the concern was raised by Lefohn and Foley (1992) and recently noted by Musselman et al. (2006) in a critical review of the use of exposure- and flux-based indices in assessing vegetation effects, the Staff Paper (EPA, 2007) ignored this important consideration except for describing the concern. On page 8-20, the authors state, “Specifically, a few commenters (sic) were of the view that a W126 (or SUM06) was not sufficient in and of itself but should be combined with a measure of the number of peaks above 100 ppb (N100).”
An example of what happens when only the W126 cumulative exposure index is used to predict vegetation effects is discussed in the following pages. Figure 12 (from page 7-28 of the
22
Staff Paper) illustrates the estimated 12-hour, 3-month W126 cumulative exposure for 2001 for the United States. Adjusting the hourly average concentrations downward by 10% resulted in applying “corrected” W126 values to estimate black cherry biomass loss. Figure 13 (from page 7H-3 in the Appendix to the Staff Paper) illustrates the predicted losses for black cherry in 2001. Between 20 to 30% biomass loss was estimated for black cherry for 2001 for the Allegheny National Forest in Pennsylvania.
Figure 12. 12-hour, 3-month W126 cumulative exposure for 2001. Source: EPA (2007).
Unfortunately, there is little or no ground truth associated with the biomass loss predictions and thus, EPA is unable to confirm any of the loss estimates. Thus, at best, the exposure-response models used in the Staff Paper (EPA, 2007) are theoretical models used to “estimate” vegetation effects; information about the uncertainty associated with the predictions is lacking. Given the large biomass loss estimated for black cherry for 2001, I have reviewed the U.S. Department of Agriculture Forest Service’s bioindicator surveys to explore whether some indication exists for large biomass losses for black cherry for the 2001 growth season.
Morin et al. (2006) reported on the O3 bioindicator survey results conducted for the Allegheny National Forest for the period 1998 and 2001 for both black cherry and pin cherry for each year. The amount of symptoms exhibited (an indicator of a physiological response) on the black cherry was less for the years when the Palmer Hydrologic Index was the lowest (i.e., 1999 and 2001) (http://www1.ncdc.noaa.gov/pub/data/cirs/). Morin et al. (2006) also reported survey results from 173 plots on the Allegheny National Forest. The authors recorded observations on the health of black cherry trees between 1998 and 2001. Observations were made of the following forest health indicators: crown dieback, crown density, crown ratio, foliage transparency, the types of tree damage, and the amount of standing dead trees. None of the results for the forest health indicators showed any concerns for the health of the population of black cherry in the Allegheny National Forest. If O3 exposures were significantly impacting the carbohydrate production or causing the trees to utilize carbohydrate reserves, there could be an increase in the amount of crown dieback or mortality. Some crown dieback was observed on black cherry, but this was associated with areas which were defoliated by cherry scallopshell moth and/or elm spanworm. Mortality of black cherry was greatest in the smallest diameter class, which the authors reported “... likely was due to self-thinning of this shade-intolerant species.”
List of Depositions, by party to the lawsuit, he is listed as FSCL (Florida Sugar Cane League - hint, they aren't an environment NPO) witness. In the actual deposition, at 311/6, he is virtually accused of dishonesty in reviewing the data
MR. NETTLETON: Or whatever he's done to
7 manipulate the data from the report.
25 Q. Under 1988, this is under the heading
from this deposition it turns out he has a long, early history with the EPA, and with mining interests in Montana
31
1 Honors and Appointments, there is a reference to the
2 National Vegetation Survey/Forest Response Program.
3 Can you tell me what that is?
4 A. The National Vegetation Survey was a
5 project that the USDA Forest Service ran that had
6 deposition, wet deposition, ozone, sulfur dioxide,
7 nitrogen dioxide, meteorology, forest resources,
8 maybe even forest pest data in the data set, data
9 base.
10 And this was a project that Dr. Don Marks
11 of the US Forest Service out of Georgia ran and I
12 participated in that program, characterizing wet
13 deposition and air quality for the USDA Forest
14 Service. And as a result of that in 1988 I received
15 a certificate of appreciation.
16 Q. Can you tell me, first of all, what was
17 the survey about? What was the purpose of it?
18 A. Essentially they are looking at forest
19 health. And the concern was that the trees were
20 having reduced growth and they wanted to know whether
21 that was associated with natural causes or man-made
22 causes.
23 And it was not a lot of people going out
24 and gathering new data but actually the summarization
25 of existing data. To the best of my knowledge, I
32
1 don't recall, I don't recall but I think most of it
2 had to do with looking at what data was available and
3 characterizing that and putting it in a form to do
4 overlays of data sets to look for cause and effect
5 relationships.
6 Q. Do you recall what the conclusions were of
7 that survey?
8 A. I don't think there are conclusions. I
9 think it is a data set at this point.
10 Q. What was your involvement with that
11 survey?
12 A. I think I just indicated that, which was
13 to characterize the gaseous and the wet deposition
14 data.
15 Q. You didn't do any analysis to determine
16 cause and effect in that regard?
17 A. I don't think there was cause and effect.
18 I don't think there was effect type part, I think it
19 was the characterization of the data as they existed
20 but not the overlay as to here are effects, here's
21 what air quality was observed, what can we best say
22 what is causing that. To the best of my knowledge
23 they never put together a team to look at that.
24 However, it was indicated to me that when
25 they did, if they did one, they desired me to
I was asked by the EMAP program to write a
3 chapter or a section dealing with how to characterize
4 exposure in deposition data in potentially biological
5 meaningful terms that could relate to these indices
6 if and when there were changes in the indices, and
7 the flag went up, yellow or red flag, whatever,
8 saying changes are occurring here, what could they
9 be. And so I wrote that chapter for them.
10 Q. When you are talking about
11 characterization, what do you mean?
12 A. In the area of air quality for, let's say,
13 sulfur dioxide you measure 8,760 hourly values this
14 year. You have an individual way you can take those
15 hourly measurements, one way to is to take a simple
16 arithmetic mean and you smear the information at that
17 point, meaning you just put it all in one pot and say
18 there is one average.
19 In the area of research that I have been
20 involved in, the higher concentrations are more
21 important than the lower concentrations so the
22 mathematics of what we have been doing have been
23 associated with how to design the indices that we use
24 for exposure such that they focus on the higher
25 concentrations. So for us using simple arithmetic
36
1 means are inappropriate.
2 Q. Are appropriate?
3 A. Inappropriate. And the discussion of that
4 chapter involved why you focus on the peaks, the peak
5 concentrations and how you take the data and put it
6 in a form that gives the researchers who worry about
7 the effects part a better handle on what the exposure
8 looks like.
9 Q. Was there a particular geographic area
10 that was involved in this mapping?
11 A. No. It is a national program. Well, the
12 EMAP program is not a mapping program. It is a
13 national program in scope. What they asked me to do
14 was write a general chapter.
15 Q. Were there any types of conclusions that
16 resulted from your work in this chapter?
17 A. The conclusion that I drew is that based
18 on the higher concentrations being more important
19 than the lower concentrations from a biological
20 perspective, one would characterize the air quality
21 and deposition in the particular form that I
22 discussed in the chapter. So they were bottom line
23 conclusions, yes.
24 Q. Am I correct that your work in this regard
25 did not relate to any specific site that you were
37
1 trying to analyze?
2 A. That is correct.
3 Q. What chemical parameters were you dealing
4 with in that chapter?
5 A. I know for sure we were dealing with
6 ozone.
he lawsuit, and it was also quite instructive - natural forces are to blame - the exact same thing he is still saying about ozone! He worked out his maThe middle of page 7, you have a report
24 entitled Characterizing Mountain Cloud Chemistry Data
25 in Support of the Spruce-Fir Research Cooperative
39
1 Effort.
2 A. Yes.
3 Q. Can you tell me what that was about?
4 A. Yes. That was a short report.
5 We had a meeting at the State University
6 of New York in 1988, I believe, that resulted in
7 about 15 scientists from across the United States
8 meeting and talking about the best way of handling
9 mountain cloud chemistry data.
10 The idea there is that they were gathering
11 cloud information, wet deposition information and
12 ozone and nitrogen dioxide information and how best
13 to put the data into a form that would allow once
14 again for cause and effect analysis.
15 And I, of course, played a key role in the
16 characterization of those data. At that point we
17 were not analyzing the data, we were talking about
18 what to do with the data. That was a short report.
19 Q. So you didn't reach any conclusions as to
20 the cause and effect?
21 A. No. Once again we focused on the peaks
22 being important.
mes long ago and simply repeats them...peaks are more important and predictive than background averages - this is like the disbersants in the gulf for the oil - high smokestacks...and it's naturally occurring anyway...and the damage isn't that bad...15 Q. The next report listed there is Natural
16 Processes and Their Possible Importance on the
17 Chemistry of Selected Lakes in New York and
18 Massachusetts. Can you tell me what that was about?
19 A. The coauthor with me on that report was
20 Dr. Ed Krug. We were looking at the possible
21 importance of the soils and vegetation in
22 contributing to the acidification of lakes that
23 previously had been thought to be, acidification
24 thought to be caused directly by acid deposition.
25 What we covered in that report was the
38
1 various processes, soil processes that could be
2 occurring that would result in acid going off into
3 the lakes or streams.
4 Q. What conclusions did you reach?
5 A. A very important conclusion, that is that
6 the natural processes were contributing to the
7 acidification of lakes and streams.
8 And our work essentially which
9 complemented the work that Dr. Krug had published in
10 1983 and I published in 1985, our work ended up being
11 the driving force for the National Acid Precipitation
12 Assessment Program concluding that natural processes
13 do play very important roles in the acidification of
14 lakes.
he this part is chilling...
The middle of page 7, you have a report
24 entitled Characterizing Mountain Cloud Chemistry Data
25 in Support of the Spruce-Fir Research Cooperative
39
1 Effort.
2 A. Yes.
3 Q. Can you tell me what that was about?
4 A. Yes. That was a short report.
5 We had a meeting at the State University
6 of New York in 1988, I believe, that resulted in
7 about 15 scientists from across the United States
8 meeting and talking about the best way of handling
9 mountain cloud chemistry data.
10 The idea there is that they were gathering
11 cloud information, wet deposition information and
12 ozone and nitrogen dioxide information and how best
13 to put the data into a form that would allow once
14 again for cause and effect analysis.
15 And I, of course, played a key role in the
16 characterization of those data. At that point we
17 were not analyzing the data, we were talking about
18 what to do with the data. That was a short report.
19 Q. So you didn't reach any conclusions as to
20 the cause and effect?
21 A. No. Once again we focused on the peaks
22 being important.
The Electric Power Research
23 Institute in Palo Alto, California, asked me to
24 review the various models that were being used in the
25 acid rain area.
next we hear how the EPA has asked him to "look into" things...it would be an interesting excercise to do a chart as to under which administration EPA was asking him to "look into" things and under which administrations he started writing complaints about regulations.
13 Q. On page 10, the second listed report, A
14 Review and Assessment of the Effects of Pollutant
15 Mixtures on Vegetation Research Recommendations.
16 Can you tell me what that report is about?
17 A. Yes. EPA undertook a very novel exercise.
18 About 1983 they came to me and asked how much
19 information was available concerning what was known
20 about the combination of sulfur dioxide, ozone and
21 nitrogen dioxide in the atmosphere, because prior to
22 that time exposure experiments had been done in a
23 laboratory where plants were fumigated with what we
24 call square wave type exposures which means constant
25 concentrations. And the concern that EPA had was
45
1 they were using pollutant mixtures in the laboratory
2 that were not really observed in the real world.
3 So what EPA did, asked if I would evaluate
4 the data and draw conclusions concerning what
5 pollutant mixtures were in the atmosphere and in what
6 concentrations and when did they occur.
7 As a result of my work and work of other
8 researchers who were summarizing what was known about
9 the effect of pollutant mixtures on vegetation
10 itself, we held a workshop in 1983 and in January of
11 1984 we published the summary report. And I was
12 responsible as one of the coeditors for its
13 publication.
14 Q. What pollutant were you looking at?
15 A. Sulfur dioxide and nitrogen dioxide and
16 ozone.
17 Q. Was that all?
18 A. Yes.
19 Q. What conclusions did you reach?
20 A. That the EPA was using, or not just EPA
21 but the research activities that were being conducted
22 were using types of exposures that would be
23 inappropriate in helping to set a standard. In other
24 words, the data themselves could help discuss the
25 mechanisms that were involved in the effects that
46
1 were observed but might not be very much help in
2 helping to set the level and form of the standard to
3 protect against such mixtures.
m...pea
In 1983 I was asked by the American
14 Petroleum Institute to review the major
15 accomplishments and some of the weaknesses associated
16 with the EPA's National Crop Loss Assessment Network.
17 And my role in 1983 was essentially to work with my
18 peers, some of whom were participating in that
19 program, to essentially -- I was then responsible for
20 drawing conclusions based on the information, written
21 information that I had received from my peers as well
22 as from the government.
23 That was the report in May of 1983. By
24 1985 I had been asked by the United States
25 Environmental Protection Agency to actually join the
50
1 National Crop Loss Assessment Network based on the
2 critique that I had done of the program in that my
3 major emphasis there was that they, I felt very
4 strongly they were not properly handling some of the
5 air quality characterization information in an
6 appropriate way.
7 And its response to me, meaning the
8 agency, was to invite me to participate by reviewing
9 with Dr. Jennifer Logan and several others the
10 Kriging techniques that EPA was using. I believe it
11 was '84 and '85, we were asked to do that.
12 And there were five of us that reviewed
13 and evaluated Kriging technique. This information
14 was very important because it was feeding directly
15 into the economic assessment loss estimates for the
16 nation.
17 Q. I believe you said that you criticized the
18 way that EPA was handling --
19 A. Critiqued.
20 Q. Critiqued. Can you tell me what the
21 nature of your critique was?
22 A. Yes. It later was supported in the
23 criteria document of 1986 by the United States EPA
24 for the criteria document for ozone, and that is that
25 the agency was characterizing its ozone data in 7
51
1 hour average information.
2 What that meant is they were looking only
3 at a time period 9:00 in the morning until 1600 in
4 the afternoon, 4:00 p.m. in the afternoon, and
5 averaging that information.
6 The criticism was that, and it was
7 constructive criticism, was that with peaks being
8 important, the high concentrations being important,
9 that they were averaging out the information and that
10 ultimately for the same seven hour average over a
11 growth season you could end up with some exposure
12 regimes with peaks and some without.
13 EPA in 1986 agreed and concluded that the
14 long-term average was an inappropriate way to
15 characterize ozone exposure for vegetation effects
16 research.
17 Q. Why is it in your view the peaks that are
18 the important aspect in the effects of whatever
19 parameter you are looking at?
20 A. Well, for vegetation.
21 Q. Okay, for vegetation effects.
22 A. That is not a general global statement.
23 For vegetation, it had to be based on the
24 biological experiments. That was the key to it.
25 And I helped design the exposure regimes
52
1 in 1983 with the US EPA in Corvallis that essentially
2 had some exposure regimes that had peaks in them and
3 some that didn't but with the same seven hour mean.
4 What EPA found was that in fact even with
5 the same 7 hour mean they had more damage, more
6 growth reduction on their agricultural crops with
7 those exposures that had the peaks in them.
8 Q. With regard to the research you have done
9 in this regard, what specific chemicals or compound
10 or what have you determined the peaks are that are
11 the important aspect of the effect on vegetation?
12 A. Sulfur dioxide, ozone and possibly
13 nitrogen dioxide. Those are the three major
14 pollutants that are in the national ambient air
15 quality standards.
16 Q. Have you conducted any similar research to
17 formulate any opinions concerning the effects of
18 peaks in nutrient concentrations on vegetation?
19 A. Not nutrients, but pH was one that we were
20 looking at, hydrogen ion.
21 Q. I am sorry?
22 A. Hydrogen ion.
23 Q. Is it surprising that the peaks in the
24 three compounds that you referenced had a cause and
25 effect relationship?
53
1 A. Well, there were two types of exposure,
2 there was chronic which is long-term exposure and
3 acute which is a short-term concentration.
4 Certainly from our standpoint up into the
5 1980s there was nothing in the literature to our
6 knowledge that indicated for growth reduction, for
7 damage that peaks could be solely attributable to the
8 effects. It was only after we designed the
9 experiments, ran them and published it that we had
10 the definitive stuff, material that we could then
11 cite in the literature to draw the conclusions that
12 we drew.
13 Would it be a surprise? No. But the
14 bottom line on that there is as a scientist you can't
15 say I think and then run over the hill so it was
16 necessary to design the experiments and then
17 implement them and draw conclusions.
For instance, he showed up at this meeting of the California Air Resources Board 2005 Air Quality Advisory Committee Meeting...why?? WHO was paying him??
and had this to say:
For the last 25, 30 years my focus of research has been on the effects of ozone and how to characterize hourly average concentrations in a way that are biologically meaningful for vegetation. In 1980 our work led us to publish many, many papers beginning in '80 on the importance of the peaks for vegetation and how to characterize it in the form of exposure metrics that focus on those peaks and the distributions.
In 1987, Dr. Milan Hazucha from UNC published a PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
2223 1 paper. And EPA immediately called me up because the paper
2 showed that again peaks were important but this time for 3 human health, and a lot of the research I had done earlier 4 on vegetation was relevant for the human health area. 5 I'm responsible for synthesizing the various 6 ozone profiles for some of the clinical human health 7 chamber studies that are in the staff report here, in 8 particular working with Dr. Adams in his research and such 9 at UC Davis.
10 I had published over 150 peer review papers and 11 technical reports and was an executive editor of AE, 12 Atmospheric Environment, from 1989 to 1999. 13 Next please.
14 15 MR. LEFOHN: You 16 material that we supplied 17 I've worked with for the last of six years, from the 18 Department of Statistics at Stanford University. And he 19 specifically focused on the mortality and epidemiology, 20 but I think a lot of that is relevant in terms of other 21 biological endpoints. 22 And his conclusion was that the Epi evidence 23 cannot be used to draw robust conclusions regarding the 24 circumstances and magnitudes of ambient ozone mortality, 25 in particular whether reported ozone effects are
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
--o0o-- have a very large amount of you, that Dr. Paul Switzer who
24
1 causative. Without a clear understanding of the reasons 2 for inconsistent effect estimates, one cannot rule out the 3 possibility -- and this is very important -- that the 4 ozone effect estimates are null artifacts. 5 Next please. 6 --o0o-- 7 MR. LEFOHN: And recently, as pointed out a few 8 minutes ago, the Bell, et al., piece was published in 9 JAMA. Those are -- that's the ordering of the 95 cities,
10 with the red on the left being Hawaii. And the greatest 11 road of risk per 10 ppb is at the bottom, and the least is 12 at the top. And Los Angeles is very near the top. And to 13 the right is the eight-hour average design values that 14 I've determined for the 2001 to 2003 time period. And 15 obviously some of the things that you've all been talking 16 about other things such as associate demographic and other 17 pollutants, et cetera, certainly could be impacting what's 18 going on here. 19 But also notice that the confidence intervals are 20 going through zero for many, many, many, many of those 95 21 cities. Now, at the very bottom there's a national 22 average. So it is significant that simply because you're 23 taking a lot of nonsignificant numbers and -- you have a 24 lot and when you divide, you get significance. It doesn't 25 mean that that is a meaningful number.
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
1 2 3 4 5 6 7 8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Next over here please. --o0o--
MR. LEFOHN: These are the numbers from Honolulu, Hawaii, for the eight-hour daily maximum concentration from 1991 to 2000. Please note they're very, very low. So the question here: Do we need to pay attention to the magnitude, because it was done for the entire time period.
Now, the interesting thing is, most -- we talk about season versus year. But in fact the way that EPA does report its data from the majority of cities across the United States, it is for the ozone season. In other words you're not going to find data, except for California and a few other states, for 12 months. And so you have seasonal data essentially for many of the places.
But this is a very low eight-hour daily maximum. It actually -- this city was the lowest of the 95 cities.
So there's certainly reasons one can say there are confounding influences, there are all kinds of other things. We still believe in the epidemiological results. It's just they're highly uncertain.
One other aspect brought up is that you're
Next please.
which Professor Switzer has looking at a modeling artifact.
--o0o-- MR. LEFOHN: Stratospheric ozone over Honolulu,
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
25
26 1 Hawaii, the elevation is three meters. On March 9th-10th
2 2004, NOAA flew over the area as part of research project. 3 And I received the telephone call probably around that 4 time asking if I had data for Honolulu, because they 5 actually were tracking stratospheric intrusion and
6 wondered how far it got to the -- if it got to the 7 surface. It was seen at Monoloa over a 100 parts per 8 billion. Certainly that is a trackable concentration at 9 the high elevation Monoloa site. But down near the ground
10 where you're not seeing over 100 ppb but in the 50's and 11 60's it's certainly a lot different. 12 Those are the two days which usually -- some 13 people are tornado chasers. I'm a stratospheric ozone 14 chaser. And so one of the things I look for when I look 15 at surface are enhanced levels. But enhanced levels are 16 not 100 parts per billion. They're 40 or 50 parts per 17 billion, but constant, meaning you have two days', three 18 days' worth without any break in the evening or the
19 morning. 20 Next please. 21 --o0o-- 22 MR. LEFOHN: One of the things that one is using 23 in the estimates for epidemiology is the center-city-type 24 analysis, where you basically take all the monitors within 25 a county and you average it, and you basically say, "This
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
27 1 is my number that I'm going to use for that county and I'm
2 going to use that on a daily basis" in whether it's 3 8-hour, 1-hour daily max or 24 hour. 4 In order to make that work, with a linear model, 5 you only need that the correlation coefficients are high. 6 If you're dealing with a non-linear situation -- which Dr. 7 Switzer argues that there is indications for 8 non-linearity -- and in fact the JAMA paper with their 9 results argues for it also -- is that the absolute
10 concentration has to be small too among monitors. This 11 analysis I did as part of the criteria document that -- we 12 did it for 24 areas across the United States. And I have 13 segregated the data for California. What it says is that 14 the minimum correlation coefficient is fairly low for most 15 of the areas that we're looking at. And low, while I know 16 is a subjective thing, but the point is it's not .9, .8; 17 it's .2, .3, .4, et cetera. 18 The max correlation of course is fairly high. 19 But you have a range of correlations depending upon which 20 days the data are among the pairs. In addition, on the 21 right side, the last two columns, are the minimum P90's, 22 which is the 90th percentiles of the differences of 23 absolute concentrations. And all we're doing is arranging 24 those in a percentile distribution and picking the 90th 25 percentile. What it says is the minimum 90th percentile
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
28 1 runs fairly large actually. And the max is fairly large,
2 40, 50 parts per billion many times. 3 So what we're seeing here is that the correlation 4 coefficients are not real high and the absolute 5 concentration values are not real low in terms of the 6 differences. In other words, some sites are going up and 7 other sites are not necessarily moving in the same amount 8 or same direction all the time. In addition, the absolute 9 concentrations, they're not zero. So if you have a linear
10 model, the correlation coefficient is not helping you 11 achieve the assumptions that you've put in it. If you 12 have a non-linear model the absolute concentrations become 13 important and where the people are becomes important 14 within the cities you're looking at. 15 Next please. 16 --o0o-- 17 MR. LEFOHN: A bottom-line concern about the use 18 of the Epi data in the standard-setting process, at this 19 time Epi results, my recommendation, should not be used to 20 establish either the level or the margin of safety for the 21 ozone standard. There's too much variability. As I've 22 just showed you, some of the assumptions are just not 23 being met, which may explain the heterogeneity among 24 cities that we're seen for the 95 city study as well as we 25 saw for PM.
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
1 2 3 4 5 6 7 8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
29
The available epidemiologic evidence on ozone mortality cannot be used to draw robust conclusions regarding causation. And the ozone standard, my recommendation, should be based on results associated with human exposure chamber studies that apply realistic exposure patterns. And what I mean by realistic is the eight-hour square waves that were used at constant concentrations are rarely found in the United States, rarely.
Next please.
--o0o-- MR. LEFOHN: Policy-relevant background. This is
a direct quote right from the document itself. I want to differentiate between natural background and policy-relevant background. It's important. Within the range of concentrations due to such external or controllable sources those concentrations that may impact determinations of compliance with air quality standards or limit the potential for air quality improvements due to control programs have been defined in the document as policy-relevant background. This is not the same as natural background, as I just said.
Next please.
--o0o-- MR. LEFOHN: There's a large variability among
PETERS SHORTHAND REPORTING CORPORATION (916) 362-234530
1 global models on the attribution of the contribution of 2 natural ozone to policy-relevant background. One global 3 model that staff has focused on, which is the Fiore, et 4 al., model, estimates that natural background ozone levels 5 in four-hour afternoon average concentrations -- those are 6 not hourly averages -- are in the 10 to 25 ppb range and 7 never exceed 40 ppb. In other words, natural background 8 will never exceed 40 ppb, never, never. 9 Next please.
10 --o0o-- 11 MR. LEFOHN: Staff states in its December 3rd, 12 2004, response to the comments: "Data on 19th century 13 ozone concentrations measured in Europe and the U.S. 14 (Bojkov, '86) show that spring peak ozone partial 15 pressures were about essentially 30 to 50 parts per 16 billion in the Midwestern U.S. and ranged from around 20 17 to 30 ppb in Europe." 18 Next please. 19 --o0o-- 20 MR. LEFOHN: I might mention that the Bojkov 21 piece has been used over and over again to state that 22 ozone was very, very low in the 19th century and was 19 23 parts per billion in Europe. Why 19 compared to the 24 numbers that I just quoted? That was the annual average. 25 Now we're dealing with daily max. But the daily maxes
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
1 2 3 4 5 6 7 8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
31 that Bojkov used were from Linhof. And Linhof had 7-hour
average values. They were daytime, 0700 to 1400, and nighttime, 2100 to 0700.
Next please.
--o0o-- MR. LEFOHN: Therefore, if the 7-hour maximum
average concentrations were in the 30 to 50 ppb range in the spring time, during pretty industrial times, the hourly average concentrations from 1871 to 1903 in Michigan had to be higher or equal to or greater, let's say, than 50 parts per billion.
Thus natural background, not policy level background, concentration levels appear to be higher than 50. The estimates for the range of policy-relevant backgrounds of course have to be greater than the background values estimated by the model sighted by staff. Once again, Fiore, et al., said background never exceeds -- never -- 40 parts per billion.
Next please.
--o0o-- MR. LEFOHN: The removal of all anthropogenic
emissions. One of my comments was: "In some of the modeling efforts to estimate natural background ozone concentrations within North America investigators removed all anthropogenic emissions of NOx, CO, and non-methane
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
32
1 hydrocarbons, including NOx emitted from aircraft and 2 fertilizer, but not biomass burning." 3 ARB's response was: "ARB does not propose a 4 projected all anthropogenic sources of ozone precursors in 5 California could be eliminated." In other words,
6 fertilizer adds to the amount of policy-relevant 7 background. 8 Next please. 9
--o0o-- averaging time's important. And that in the document itself, the
10 MR. LEFOHN: The 11 the point I was making is 12 15 to 35 parts per billion were probably long-term 13 averages. And that if you're talking about the 14 variability and the distribution and such, it's going to 15 be higher than that. And staff's response is: "We agree 16 that reading the long-term mean values presented in the 17 staff paper as absolute maxima could be misleading." And 18 it's going to be changed. 19 Next please. 20 --o0o-- 21 MR. LEFOHN: Bottom line. Because it appears the 22 policy-relevant background levels are higher than 40 parts 23 per billion assumed by staff, therefore the rollback 24 estimates for the ozone concentrations are too optimistic 25 and actually will occur slower than predicted. In the
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
1 2 3 4 5 6 7 8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
33 model that Barry Larson and his staff have worked on, the
higher the policy-relevant background, the slower the reduction will be in the mid-level concentrations.
The peaks will come down, but the rate of decline of the mid-levels are going to slow down in a drastic fashion, depending upon how high the policy-relevant background is above the assumed 40. A higher policy-relevant background will result in a greater slowing down, as I've said, in the mid-level. And, consequently, health benefits estimated by staff probably have been overestimated.
Next please.
--o0o-- MR. LEFOHN: Human health effects. Experimental
exposures of human volunteers to air pollutants under realistic varying exposure-controlled laboratory conditions have provided important information directly relevant to standard setting.
The important ramifications reported by Hazucha, et al., and Adams, et al., is that a non-linear dose response relationship is evident. This is similar to the research that I did 25 years ago on vegetation.
Next please.
--o0o-- MR. LEFOHN: The higher hourly average
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
34
1 concentrations elicit a greater effect than the lower 2 values in a non-linear manner. A major implication of a 3 non-linear dose response relationship is at the same 4 8-hour average. It's the same 8-hour average, with 5 different distributions of hourly average concentrations, 6 will elicit a different adverse effect. 7 Now, I wrote a paper in 1993 that said you 8 shouldn't use an 8-hour average because you're going to 9 get inconsistent results. But prior to that I had written
10 a paper that said if EPA continued to push for a 7-hour 11 seasonal average for vegetation, the agency would in fact 12 get inconsistent results with the same long-term mean 13 giving different effects because the distributions are 14 different. 15 Case in point, Colstrip, Montana, in the middle 16 of nowhere in eastern Montana had a 7-hour seasonal of 43 17 parts per billion. Chicago, illinois, at the time -- or 18 just outside Chicago in the county had 43 parts per 19 billion. Same average, different distribution. Chicago 20 had a lot more peaks. 21 Next please. 22 --o0o-- 23 MR. LEFOHN: Okay. For those that are interested 24 in ozone from Asia, work that I'm doing now with NOAA, 25 which is different than the aircraft stock, deals with
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
1 looking at sites from all over the world, remote sites, 2 and asking the question: Are we getting increases in 3 ozone in the southern hemisphere and the northern 4 hemisphere? The Lawson Volcanic National Park is
5 certainly not one of the pristine sites that I would 6 normally pick for the study. However, because the focus 7 of Jaffe, et al., and others have been on the Lawson 8 Volcanic National Park, I carried that along in the study. 9 And the bottom line here is that you are seeing
10 changes in the distributions. And these are showing the 11 bins, the 10 ppb bins by month over the period 1988 to 12 2003. And the negative numbers are simply saying that 13 you're losing low values because they're going up. In 14 other words you have conservation of the concentrations. 15 Now, some are going to go down from the top. And bins 16 obviously at the bottom have to go up. You have a
17 different process here of low end coming up, the bottom 18 line being that it's not -- there were spring trends, but 19 the spring trends are not in March. They're in April, 20 May; and then summer, you get June and July. 21 The bottom line from what I'm seeing from the 22 data is that, yes, indeed you are seeing trends during the 23 period, which I agree with the authors. However, you have 24 the Redding, California, and the, Anderson, California 25 site that also are subject to very high levels of ozone
PETERS SHORTHAND REPORTING CORPORATION (916) 362-2345
35
36
1 starting in April. And whereas the authors, Jaffe, et 2 al., said that they're sure that Lawson has seen Asia 3 because ozone does not form in high concentrations in the 4 spring time, I think that's incorrect based on my 5 experience of looking at your data.
6 7 8 9
10 11 12 13 14 that set 15
Next. Is that it? We're done. Thank you. HEALTH AND EXPOSURE ASSESSMENT BRANCH CHIEF BODE: John Heuss. (Thereupon an overhead presentation was Presented as follows.) MR. HEUSS: Thank you all. She's trying to get up. I am John Heuss with Air Improvement Resource, a
16 consultant to the Alliance of Automobile Manufacturers. 17 I'd like to thank you for the opportunity to 18 provide these comments.
19 20 comments 21 I'm sure 22 night you had a chance to read it. Would certainly ask 23 that you do read it and consider it before you come to 24 closure on the document. 25 --o0o--
A doctor testified:
Recent studies have also suggested that ozone 9 exposure in children may impair long-term lung function
10 and may not only exacerbate asthma, but actually cause it. 11 So, in summary, I urge you to adopt the new 12 8-hour average of 70 parts per billion standard not to be 13 exceeded and the one hour average of 90 parts per billion 14 not to be exceeded.
The Western States Petroleum Association (WSPA) and the American Petroleum Association (API) represent companies that explore, develop, refine, market and distribute petroleum and petroleum products throughout the United States. WSPA is composed of nearly 30 companies that have operations within the 6 Western States including California. API is the national trade association of America's oil and natural gas industry, and represents more than 400 members involved in all aspects of the oil and natural gas industry. Both WSPA and API members are owners and operators of major facilities regulated under the California ozone standard, and producers and marketers of fuels that are often targeted as a means to reduce ozone precursors. As such, we have a direct and substantial stake on the outcome of this proposal. With this letter, we are providing comments on the draft benefits analysis (Chapter 10 of the Review of the California Ambient Air Quality Standard for Ozone) that has been developed by the California Air Resources Board.
We have asked Dr. Allen Lefohn of ASL and Associates and Dr. Stanley Hayes of Environ International Corporation to help us review Chapter 10 and prepare detailed comments. These nationally recognized experts have been deeply involved in many aspects of ozone research, have authored important sections of current and past chapters for the federal Ozone Criteria Documents or conducted research cited in previous Ozone Staff Papers, and helped develop and apply approaches for assessing ozone risk
Our review has found significant problems in the methodology and assumptions that are used in the benefits assessment. We feel that the document should not be finalized until these problems are resolved This chapter moves far beyond previous health risk assessments conducted by EPA in their 1997 Ozone Staff paper, and introduces the concept of a link between ozone and mortality, which is likely an artifact of the analysis methodology and statistics. As such, we recommend that any benefits estimates predicated on mortality be removed, as these are not scientifically supportable.
It is critical that Staff address the issues raised herein, and a second draft of the Staff Paper be circulated for public review and comment. The most significant problems with the benefits analysis fall into the two areas outlined below. These problems are also described in more detail in our attached comments.
1. Inappropriate Choice of Rollback Model and Background Ozone Level
One of the most important aspects of the O3 health benefits analysis is the adequacy of the rollback methodology. If the methodology is not adequate, then the application of the rollback model will introduce a level of uncertainty that may make the health benefits analysis unreliable. We believe this has happened. Staff has underestimated policy-relevant background at many areas across California by subjectively selecting a value of 0.04 ppm. Our analysis shows that the hourly average concentrations of policy-relevant background near sea level and at higher elevations will be higher at many locations than estimated by Staff.
In addition Staff concluded that the rate of change in the concentrations above background was similar among the percentiles and this observation justified its application of a constant percentage rollback to all sites within an air basin. In carefully reviewing the information in the Appendix for the rate of change in the concentrations above background from the 1980s, the data do not show a constant percentage change above background for each of the percentiles for the sites provided in the Appendix. Also, the rollback model used in the health benefits analysis does not necessarily mimic the slowing down of the mid-level hourly average concentrations in relation to the higher hourly average values and therefore, the predicted distribution of the concentrations may not be reliable.
2. Accrual of Health Benefits Below the Proposed Standards
The proposed California standards are based primarily on chamber studies involving human subjects. The chamber studies do not show statistically significant effects below the 8-hour, 0.08 ppm level. This led Staff to propose a standard of 0.07 ppm, to include a margin of safety. Note that in our prior comments, we cited research
1115 11th Street, Suite 150, Sacramento, California 95814 (916) 444-9981 • FAX: (916) 444-8997 • www.wspa.orgDr. Drechsler – P. 3
by Dr. William Adams from UC Davis that does not support Staff’s recommendations for a 0.07 ppm standard. This was based on analyses reflecting the inappropriateness of Staff choices for background level and the use of averaging exposures for the form of multi-hour standard.
In Chapter 10 no rationale is provided for accumulating health benefits below the level of the standard to an assumed policy-relevant background level of 0.04 ppm. Since approximately 76% - 86% of the benefits presented in Chapter 10 are accrued at O3 concentrations between the proposed standards (0.07 ppm 8-hour and 0.09 ppm 1- hour) and background, accumulating benefits below the level of the standard enormously exaggerates the benefit of the proposed standard.
In addition, for more than twenty years, response to ozone has been directly measured in controlled human chamber studies. Ozone exposures in those studies have been characterized, e xercise levels have been specified, controls to avoid confounding factors have been applied, and response indicators have been measured. Collectively, this body of human chamber studies provides a large and robust database with which to characterize human response to ozone over a wide range of ozone levels and lung function, respiratory symptom, and other endpoints.
However, none of these data are used in the benefits assessment in Chapter 10. Instead, Chapter 10 relies entirely on the results of epidemiological studies, bypassing data from human chamber studies. We strongly recommend that Chapter 10 be revised to include those chamber data in its assessment and to reconcile those data with assumptions made in the chapter based on epidemiological studies
Staff appears to have accepted epidemiological studies, many done for other purposes (e.g., PM) as being sufficient to establish causality and quantify concentration-response relationships for ozone. In fact, the large uncertainties and inconsistencies in this literature, documented by Staff and us, preclude interpreting and using the current epidemiological data in this fashion – particularly for ozone. For reasons not explained, the inconsistencies and model dependency of epidemiologically derived effect estimates – although recognized by Staff – are often subsequently ignored. It appears that there are greatly different standards of evidence that are being used in standard setting versus benefits estimation.
WSPA and API Recommendations
We do not believe that the benefits estimates as currently presented in Chapter 10 are based on sound scientific data. The estimates do not (1) use the correct body of data (chamber studies), (2) reflect the correct interpretation of the data used (e.g., assumes that acute mortality is causally linked to ambient ozone exposure), (3) use an appropriate estimation methodology (accrue benefits below the level of the standard and rely on linear concentration-response models), or (4) reflect the range of uncertainty associated with the data (potential unresolved confounding). As such, we fail to see
1115 11th Street, Suite 150, Sacramento, California 95814 (916) 444-9981 • FAX: (916) 444-8997 • www.wspa.org
Dr. Drechsler – P. 4
how the benefit estimates will help inform policy makers with their decisions on the proposed California ozone standard.
The chapter should be revised to address the significant issues raised in our comments before being presented to either the Air Quality Advisory Committee or the Board. In addition, to the extent that Staff or the Board feels it is necessary to have a benefits assessment performed, additional estimates should also be presented to the Board. At a minimum, benefit estimates should be based on models derived from chamber studies. One set of estimates should address the incremental benefits of achieving the proposed California ozone standard as opposed to meeting the current federal ozone standard. Estimates are also needed for the benefits accrued by meeting the proposed standard without accumulating benefits to the level of policy relevant background. All estimates based on epidemiological studies should incorporate the full range of uncertainty stemming from these studies. We recommend, however, that Staff remove any benefits estimates predicated on mortality, as these are likely artifacts of the analysis methodology.
Lastly, since the basis for the proposed California standard is identical to that used by EPA in their 1997 ozone Staff Paper – that is, human exposure effects demonstrated from chamber studies -- we believe that Staff should estimate ozone’s impact by performing a risk assessment using a methodology similar to that used in the 1997 EPA Staff Paper.
We welcome the opportunity to continue discussions with your agency. After you have had a chance to review these submittals, please feel free to contact me at 310-808-2149, Mr. Kyle Isakower (API) at 202-682-8314, or Dr. Mark Saperstein, (BP, Chair of WSPA Task Force) at 714-228-6716.
Cc: Dr. Alan Lloyd Ms. Catherine Witherspoon Mr. Mike Schieble Ms. Catherine Reheis-Boyd
1115 11th Street, Suite 150, Sacramento, California 95814
this is really scary shit!Dear Dr. Drechsler:
The Western States Petroleum Association (WSPA) and the American Petroleum Association (API) represent companies that explore, develop, refine, market and distribute petroleum and petroleum products throughout the United States. WSPA is composed of nearly 30 companies that have operations within the 6 Western States including California. API is the national trade association of America's oil and natural gas industry, and represents more than 400 members involved in all aspects of the oil and natural gas industry. Both WSPA and API members are owners and operators of major facilities regulated under the California ozone standard, and producers and marketers of fuels that are often targeted as a means to reduce ozone precursors. As such, we have a direct and substantial stake on the outcome of this proposal. With this letter, we are providing comments on the draft benefits analysis (Chapter 10 of the Review of the California Ambient Air Quality Standard for Ozone) that has been developed by the California Air Resources Board.
We have asked Dr. Allen Lefohn of ASL and Associates and Dr. Stanley Hayes of Environ International Corporation to help us review Chapter 10 and prepare detailed comments. These nationally recognized experts have been deeply involved in many aspects of ozone research, have authored important sections of current and past chapters for the federal Ozone Criteria Documents or conducted research cited in previous Ozone Staff Papers, and helped develop and apply approaches for assessing ozone risk.
1415 L Street, Suite 600, Sacramento, California 95814 (916) 444-9981 • FAX: (916) 444-5745 • www.wspa.orgDr. Drechsler – P. 2
Our review has found significant problems in the methodology and assumptions that are used in the benefits assessment. We feel that the document should not be finalized until these problems are resolved This chapter moves far beyond previous health risk assessments conducted by EPA in their 1997 Ozone Staff paper, and introduces the concept of a link between ozone and mortality, which is likely an artifact of the analysis methodology and statistics. As such, we recommend that any benefits estimates predicated on mortality be removed, as these are not scientifically supportable.
It is critical that Staff address the issues raised herein, and a second draft of the Staff Paper be circulated for public review and comment. The most significant problems with the benefits analysis fall into the two areas outlined below. These problems are also described in more detail in our attached comments.
1. Inappropriate Choice of Rollback Model and Background Ozone Level
One of the most important aspects of the O3 health benefits analysis is the adequacy of the rollback methodology. If the methodology is not adequate, then the application of the rollback model will introduce a level of uncertainty that may make the health benefits analysis unreliable. We believe this has happened. Staff has underestimated policy-relevant background at many areas across California by subjectively selecting a value of 0.04 ppm. Our analysis shows that the hourly average concentrations of policy-relevant background near sea level and at higher elevations will be higher at many locations than estimated by Staff.
In addition Staff concluded that the rate of change in the concentrations above background was similar among the percentiles and this observation justified its application of a constant percentage rollback to all sites within an air basin. In carefully reviewing the information in the Appendix for the rate of change in the concentrations above background from the 1980s, the data do not show a constant percentage change above background for each of the percentiles for the sites provided in the Appendix. Also, the rollback model used in the health benefits analysis does not necessarily mimic the slowing down of the mid-level hourly average concentrations in relation to the higher hourly average values and therefore, the predicted distribution of the concentrations may not be reliable.
2. Accrual of Health Benefits Below the Proposed Standards
The proposed California standards are based primarily on chamber studies involving human subjects. The chamber studies do not show statistically significant effects below the 8-hour, 0.08 ppm level. This led Staff to propose a standard of 0.07 ppm, to include a margin of safety. Note that in our prior comments, we cited research
1115 11th Street, Suite 150, Sacramento, California 95814 (916) 444-9981 • FAX: (916) 444-8997 • www.wspa.org
Dr. Drechsler – P. 3
by Dr. William Adams from UC Davis that does not support Staff’s recommendations for a 0.07 ppm standard. This was based on analyses reflecting the inappropriateness of Staff choices for background level and the use of averaging exposures for the form of multi-hour standard.
In Chapter 10 no rationale is provided for accumulating health benefits below the level of the standard to an assumed policy-relevant background level of 0.04 ppm. Since approximately 76% - 86% of the benefits presented in Chapter 10 are accrued at O3 concentrations between the proposed standards (0.07 ppm 8-hour and 0.09 ppm 1- hour) and background, accumulating benefits below the level of the standard enormously exaggerates the benefit of the proposed standard.
In addition, for more than twenty years, response to ozone has been directly measured in controlled human chamber studies. Ozone exposures in those studies have been characterized, e xercise levels have been specified, controls to avoid confounding factors have been applied, and response indicators have been measured. Collectively, this body of human chamber studies provides a large and robust database with which to characterize human response to ozone over a wide range of ozone levels and lung function, respiratory symptom, and other endpoints.
However, none of these data are used in the benefits assessment in Chapter 10. Instead, Chapter 10 relies entirely on the results of epidemiological studies, bypassing data from human chamber studies. We strongly recommend that Chapter 10 be revised to include those chamber data in its assessment and to reconcile those data with assumptions made in the chapter based on epidemiological studies
Staff appears to have accepted epidemiological studies, many done for other purposes (e.g., PM) as being sufficient to establish causality and quantify concentration-response relationships for ozone. In fact, the large uncertainties and inconsistencies in this literature, documented by Staff and us, preclude interpreting and using the current epidemiological data in this fashion – particularly for ozone. For reasons not explained, the inconsistencies and model dependency of epidemiologically derived effect estimates – although recognized by Staff – are often subsequently ignored. It appears that there are greatly different standards of evidence that are being used in standard setting versus benefits estimation.
WSPA and API Recommendations
We do not believe that the benefits estimates as currently presented in Chapter 10 are based on sound scientific data. The estimates do not (1) use the correct body of data (chamber studies), (2) reflect the correct interpretation of the data used (e.g., assumes that acute mortality is causally linked to ambient ozone exposure), (3) use an appropriate estimation methodology (accrue benefits below the level of the standard and rely on linear concentration-response models), or (4) reflect the range of uncertainty associated with the data (potential unresolved confounding). As such, we fail to see
1115 11th Street, Suite 150, Sacramento, California 95814 (916) 444-9981 • FAX: (916) 444-8997 • www.wspa.org
Dr. Drechsler – P. 4
how the benefit estimates will help inform policy makers with their decisions on the proposed California ozone standard.
The chapter should be revised to address the significant issues raised in our comments before being presented to either the Air Quality Advisory Committee or the Board. In addition, to the extent that Staff or the Board feels it is necessary to have a benefits assessment performed, additional estimates should also be presented to the Board. At a minimum, benefit estimates should be based on models derived from chamber studies. One set of estimates should address the incremental benefits of achieving the proposed California ozone standard as opposed to meeting the current federal ozone standard. Estimates are also needed for the benefits accrued by meeting the proposed standard without accumulating benefits to the level of policy relevant background. All estimates based on epidemiological studies should incorporate the full range of uncertainty stemming from these studies. We recommend, however, that Staff remove any benefits estimates predicated on mortality, as these are likely artifacts of the analysis methodology.
Lastly, since the basis for the proposed California standard is identical to that used by EPA in their 1997 ozone Staff Paper – that is, human exposure effects demonstrated from chamber studies -- we believe that Staff should estimate ozone’s impact by performing a risk assessment using a methodology similar to that used in the 1997 EPA Staff Paper.
We welcome the opportunity to continue discussions with your agency. After you have had a chance to review these submittals, please feel free to contact me at 310-808-2149, Mr. Kyle Isakower (API) at 202-682-8314, or Dr. Mark Saperstein, (BP, Chair of WSPA Task Force) at 714-228-6716.
Cc: Dr. Alan Lloyd Ms. Catherine Witherspoon Mr. Mike Schieble Ms. Catherine Reheis-Boyd
1115 11th Street, Suite 150, Sacramento, California 95814 (916) 444-9981 • FAX: (916) 444-8997 • www.wspa.org
Sincerely,
Michael D. Wang Manager, WSPA
Dear Gail, I think you have probably reached your own personal wit's end, and no wonder. I find myself allowing my sons latitudes simply because I fear the world they must inherit, in the same way you protect your daughter. (I suspect, however, she already is aware, and is protecting you?) Let them have fun while they can. You aren't alone in risking total burn-out. Have you read Paul Kingsnorth's essay, "Confessions of a recovering environmentalist"? If not, it might help. http://www.opendemocracy.net/paul-kingsnorth/confessions-of-recovering-environmentalist
ReplyDeleteI value what you have uncovered, terrifying as it is.
Serinde
I continue to appreciate your elegiac commentary on tree death & climate issues. It
ReplyDeletecertainly has rid me of any remaining illusions that climate change is a "someday, one day" issue. Haven't read through all of the material you just posted yet, but I understand your disgust & despair. Here we have issues with climate change deniers taking scientific agencies to court, to try & force them to alter their temperature records. Sheer madness. Some stuff below, if you haven't already read enough.
Some useful links to NO2 pollution at
http://www.youtube.com/watch?v=gwJ1h-Pz3_8&NR=1, also some good overviews of ozone
pollution at http://earthobservatory.nasa.gov/Features/GlobalTraveler/,
http://tes.jpl.nasa.gov/uploadedfiles/Eld_sci101_final_B.pdf
and monthly world maps of tropospheric ozone at
http://acdb-ext.gsfc.nasa.gov/Data_services/cloud_slice/#nd. Ozone also seems to be
implicated in other issues such as enhancing polar warming :
http://www.ametsoc.org/atmospolicy/documents/May312006_DrewShindell.pdf
It all seems to stack up in suggesting that repeated ozone exposure would cause the
tree deaths you have noticed. Recent Science Daily reports indicate very high ozone
in the Eastern US this summer as it's production from NO2 is enhanced by the high
humidity & temperatures. Perhaps exceeding a threshold level of ozone might cause the decline or sudden death of plants you are seeing.
Best Wishes, John
Thank you Serinde and John, for the links, and for the encouragement!
ReplyDeleteI was looking for a book on the shelves upstairs a couple of weeks ago, and came across a textbook about climate change, with a whole chapter on ozone effects! My daughter had highlighted sections and written notes all over it, so obviously, she knows what I am saying is correct. She just chooses to live life without factoring it into her plans, as I suppose most people do whether they are aware or not that their plans cannot be fulfilled.
I can't say I blame them.
One other thing that might be relevant to what you are seeing. The IRI data library hosts an interactive map viewer that shows NDVI (a simple index of photosynthetic activity) for 2 week periods 2000 to the present, for any area. So a year-on-year comparison might help show up plant damage, e.g. try the full link below : http://iridl.ldeo.columbia.edu/SOURCES/.USGS/.LandDAAC/.MODIS/.version_005/.SNA/.NDVI/figviewer.html?my.help=more+options&map.T.plotvalue=12-27+Jul+2005&map.Y.units=degree_north&map.Y.plotlast=40N&map.url=NDVIcolorscale+X+Y+fig%3A+colors+%3Afig&map.domain=+{+%2FNDVI+0.5+1.+plotrange+%2FT+2757.+plotvalue+X+-75.199997+-73.199997+plotrange+Y+39+40+plotrange+}&map.domainparam=+%2Fplotaxislength+432+psdef+%2Fplotborder+72+psdef+%2FXOVY+null+psdef&map.zoom=Zoom&redraw.x=19&redraw.y=17&map.Y.plotfirst=39N&map.X.plotfirst=75.2W&map.X.units=degree_east&map.X.modulus=360&map.X.plotlast=73.2W&map.NDVI.plotfirst=0.5&map.NDVI.units=unitless&map.NDVI.plotlast=1.&map.plotaxislength=432&map.plotborder=72&map.fnt=Helvetica&map.fntsze=12&map.XOVY=auto&map.color_smoothing=1&map.framelbl=framelabelstart&map.framelabeltext=&map.iftime=25&map.mftime=25&map.fftime=200
ReplyDeleteAck! John G, you must think I am much much smarter than I am! I was all night last night poring through the last links you sent!!
ReplyDeleteHi Gail, I definitely had my geek hat on when I posted those links, but I felt it might help with the background to what you are saying. I wouldn't worry about the need to be smart - I think maybe too many of us thinking we are smart is one reason we are in such trouble anyway!
ReplyDeleteYes, JohnG, but the problem is, no one with professional credibility will go near this issue without proven causality. It's the same difficulty with cancer - smoking is a pretty clear link, and geographical clusters like love canal can be attributed to a particular source - but so many others cannot be proven to be caused by any certain chemical or pollutant.
ReplyDeleteI want to understand the science so that it can be shown that ozone is causing vegetation to die. Otherwise the case to stop burning fuel and creating it rests on climate change, which is too easily ignored and lied about.
Your links were terrific. I plan to forage through them again today!