Summary: The main problem with the Assessment is the absence of dissenting views. Instead, controversial items are "papered over" semantically rather than addressed squarely. An open scientific debate should focus on the following crucial issues:

** The assertion of a long-term depletion trend for global ozone is based on inadequate data and inadequate analysis, as detailed below.

** There is no credible evidence for a long-term upward trend of ultraviolet radiation at the earth's surface.

** A fair evaluation of the recent theory and of stratospheric observations leads to the conclusion that chlorine from CFCs is not the principal factor leading to ozone destruction below 25 km, where most of the ozone is located. Water, in the form of vapor or ice particles, and sulfates in the form of aerosols may play a more important role. Such a finding could have powerful significance for policy and strongly argues for at least delaying the ban on CFC production.

** Referring to the accompanying "UNEP Environmental Effects Assessment" of Nov. 1994, the skin cancer incidence rate is overestimated by a large factor that has not been quantified but could be as much as ten. Skin cancer mortality, which enters into EPA cost-benefit calculations, is overestimated by an even larger factor.

** Keeping in mind that UV intensities are higher by 200-300% in Florida compared to New England, common-sense arguments speak against the importance of other health and environmental effects stemming from a hypothetical increase in UV-B (from ozone depletion) of only about 10%.

The detailed points are given in "chronological" order, keyed to page number of the Assessment.

1.2 Difficulties with the long-term calibration of the satellite instrument are merely mentioned but not further analyzed.

1.7 The interference between ozone and sulfur dioxide in the ozone measurements is much worse than presented in the Assessment, throwing doubt on the "ozone trend" derived from Dobson ground station data.

1.12 A problem, raised in the paper of DeMuer and DeBacker but not replied to, is that the Dobson station data have been "massaged" by Bojkov (cf pp. 1.6). His work was supposed to have been published in WMO No. 35. Where is it and who can vouch for it?

1.13 The statistical analysis of ozone depletion is problematic. As Hill and Bishop have shown in their 1988 preprint, the "trend" depends on the time interval selected for analysis. (For a summary of their work, cf Singer 1990.) Their important result, throwing doubt on the statistical analysis, did not appear in their finally published paper (Bojkov et al, JGR 1990). Their employer, Allied Signal, went from being a CFC producer to a producer of substitutes.

1.14 In examining Table 1-1, I note that only about 12 Dobson stations (out of a total of 43) go back to 1957-1959. That is not much global coverage for a long enough time. Keep in mind also that only a few of the stations are well maintained and calibrated.

1.15 The analyzed record presented in the Assessment covers 1979- 94--only 15 years--and includes the Pinatubo ozone decrease.

1.19 To eliminate Pinatubo, the Assessment confines its analysis to 1979-91--only 12 years, or barely more than one solar cycle. (That would be equivalent to extracting a climate trend from only one year's worth of temperature data)

1.20 The Assessment suggests not just a linear depletion but an acceleration! This seems highly unrealistic in view of the short time span of the record.

1.32 Balloon-borne ozone sondes show an unexplained ozone trend increase in the middle stratosphere.

1.32 Again, interference from SO2. This matter needs to be addressed squarely.

1.37 A significant and unexplained fact: Why is there no observed depletion trend in the middle stratosphere, from 25 to 35 km? (Above 35 km, the CFC-ozone theory seems to work; below 25 km ozone destruction seems to be controlled by factors other than just chlorine concentration.)

2.19 The Assessment lists methane sources, but where is the literature reference to the production of stratospheric water vapor (cf also p. xxii) and ozone effects (cf p. xix)

.... Missing is any reference to the work of Zander, published in the Journal of Atmospheric Chemistry 1987, which shows no increase in stratospheric chlorine. Nor is there a reference to a 1988 review paper by Prinn, confirming this finding.

3.20 The stratospheric abundance of BrO is given as 4-10 ppt, but there is no comparison with the more recent results of Wennberg (1994). Nor is there any indication of an upward trend, which would be expected if manmade sources are important.

4.4 Shows the important reactions with OH and HO2. Both theory (Ravishankara, Science 1994) and observations (Wennberg, Science 1994) give HOx as the most important depletor of ozone in the lower stratosphere--with chlorine chemistry important only above 25 km.

4.5 If large quantities of SO2 are put into the stratosphere, for example by Pinatubo, does the subsequent aerosol formation deplete stratospheric water vapor?

4.16 Figure 4-10 (from Rodriguez, 1994) calculates HOx as the most important catalytic loss cycle between 13 and 23 km.

4.17 Figure 4-11 (from Garcia, 1994) calculates HOx as the dominant ozone loss mechanism below 22 km.

5.6 Lists the important HOx reactions that destroy ozone.

6.6 The HOx is produced from the reaction of O(1D) with H2O and CH4. How do modelers handle the increase of CH4 and H2O ? (Note: check with Brasseur, Jackman, Kinnison)

6.15 The Assessment discusses mechanisms that can affect the ozone trend, listing the injection of NOx by high flying aircraft. What about the injection of H2O? What about the observed increase in stratospheric sulfates (Hofmann, 1990) and water vapor (Oltmans and Hofmann)?

6.29 Why should ozone content decrease when the models use lower values of methane? (cf Table 6-5 on p. 6.33)

7.1 Note the increase in greenhouse effectiveness for CH4--by a factor of 1.45.

7.27 Figure 7-8 shows modeled increases in tropospheric ozone for a 20% increase in CH4. Why are Penner's values so much greater?

7.28 Discusses in detail the increased GH effectiveness of methane.

7.15 If ozone depletion results in more tropospheric UV, increasing the production of O(1D) and therefore OH, should we not observe variability in OH tied to the large natural changes in ozone?

7.24 Assesses the impact of methane increases. Methane has increased by 100% in the last 100 years; can we see the effects?

9.3 Here starts a detailed discussion of surface UV radiation and a possible upward trend. The Assessment tries to explain away the results of Scotto (Science, 1988) which show slight decreases at 8 US locations over an 11-year period.

9.10 The Assessment reports the study by Kerr and McElroy (Science, Nov. 1993), including their improbable claim of an increasing trend of 35% per year (in winter at 300 nm). (cf Figure 9-8 on p. 9.12) In the Assessment their trend value has been changed to become 35+20%; the original publication showed no error bars. The objections of Michaels (Science 1994) are trivialized by the statement: "the statistical significance of these results has been disputed (Michaels et al, 1994) because some of the results were influenced by a few days in March 1993...".