Noting that no prior study has ever isolated the impact of locally emitted CO2 on air pollution or health, Jacobson attempts to do so via a modeling exercise where CO2 feedbacks affect "(a) local temperature and pressure gradients, stability, wind speeds, cloudiness, and gas/particle transport, (b) water evaporation rates, (c) relative humidity and particle swelling, and (d) temperature-dependent natural emissions, air chemistry, and particle microphysics," and where changes in CO2 directly impact "(e) photosynthesis and respiration rates, (f) dissolution and evaporation rates of CO2 into [and out of] the ocean, (g) weathering rates, (h) ocean pH and chemical composition, (i) sea spray pH and composition, and (j) rainwater pH and composition." And employing what he calls this "nested global-through-urban 3-D model" for "telescoping domains" that cascade down from the globe to the United States, to California, and finally to Los Angeles -- both with and without the anthropogenic CO2 emissions from the targeted domain -- Jacobson concludes that the differences in the model output can be attributed solely to the anthropogenic CO2 emissions of the domain in question.

Our criticism of this approach begins with the fact that the many complex phenomena that are modeled by Jacobson are simply not well enough known to be quantitatively defined to the precision that is needed to derive meaningful results from the small CO2 perturbations that are associated with the urban CO2 dome phenomenon. He reports, for example, that urban CO2 dome increases in California increased the population-weighted air temperature there by only 0.0063°C, while it increased the state's domain-averaged air temperature by a measly 0.00046°C. These increases, in turn, led to correspondingly small increases in near-surface O3 and PM2.5 concentrations, which were then input to statistical correlations that had been derived between these parameters and premature mortality. The final result was 51 additional ozone- plus PM2.5-related premature deaths/year attributable to anthropogenic CO2 emissions, while a similar analysis for the entire United States predicted about 770 additional such deaths.

How does this country-wide number compare with the number of deaths caused by other phenomena, including various diseases and accidents? According to data in the National Vital Statistics Report of 16 September 2002, there were slightly over 2.5 million such deaths per year in the United States, which means that the premature deaths attributed to anthropogenic CO2 emissions in the analysis of Jacobson accounted for only a miniscule 0.03% of the total. And if this result justifies the imposition of economically-severe measures to reduce mankind's CO2 emissions, what more horrendous things should be done to reduce the causes of the other 99.97% of premature deaths? ... each one of which accounts for more fatalities than anthropogenic CO2 emissions do.

Even these comparisons, however, do not tell the whole story, for Jacobson acknowledges that his "modeled pollution changes and their health impacts are uncertain." In the case of ozone, for example, he specifically states that "the relationship between ozone exposure and premature mortality is uncertain." And in the case of particulate pollution, he reports that "overall, PM2.5 increased with increasing CO2, but because of the opposing effects of temperature and water vapor on PM2.5, the net positive correlation was weak (r = 0.022) and not statistically significant (p = 0.17)."

So why was this paper ever produced? Taking a cue from the world's climate alarmists -- who love to "follow the money trail" -- we see that the primary funder of Jacobson's study was the U.S. Environmental Protection Agency, which has recently declared CO2 to be a dangerous air pollutant, and which is desperately seeking evidence to support this illogical and unsupportable finding. Enough said?

Sherwood, Keith and Craig Idso

Reference
Jacobson, M.Z. 2010. Enhancement of local air pollution by urban CO2 domes. Environmental Science & Technology: 10.1021/es903018m.