An unfortunate consequence of expertise is that when experts study a problem, they tend to strongly focus on those portions of it where their expertise can be directly applied. Though that seems natural, the undesired effect is that other aspects of the problem tend to be ignored instead of being explored, often with disastrous consequences. Think, for instance, about how pesticides would be developed if by a multidisciplinary team that included ecologists as well as chemists and MBAs.
In the process of exploring the geoengineering technique of Stratospheric Aerosol Injection (SAI) and sharing my findings with you I have found that SAI research is not exempt from this problem of exclusionary focus. The present post exposes yet another aspect, this time epidemiological. We start with the story of how I stumbled upon the issue.
The plan was to write one final post on SAI, focusing on governance. In researching the background on Mexico's excellent albeit unique decision to ban SAI, I came across a statement by the culprits who accidentally enabled Mexico's response: That their act had released less sulfur dioxide into the atmosphere than a single transcontinental airplane flight. This statement led me to wonder about the volume of sulfates introduced into the troposphere and lower stratosphere by jet aviation, and that in turn to what followed. So, just to be clear, it was not my idea to consider the health effects of SAI; I was too narrowly focused: mea culpa!
The first useful article encountered was that of Kapadia, et al. [2], Impacts of aviation fuel sulfur content on climate and human health in which it was estimated that (as of 2015) the burning of aviation fuel with 600 parts per million (ppm) sulfur content results in 3600 premature deaths annually due to lung cancer and pulmonary disease, and that reducing the sulfur content to 15 ppm would save an estimated 620 of those lives. The questions that naturally arose concerned the total volume of sulfur introduced via the use of jet fuel, and comparison of that quantity with the estimated 10 - 20 million metric tons of sulfates required by SAI for effective temporary reversal of global heating. Based on worldwide jet fuel consumption for 2018 as published by Thomson-Reuters [3], 600 ppm corresponds to about 142 thousand tons of sulfates, and 15 ppm corresponds to 3540 tons. As these figures pale in comparison with the amounts needed for SAI, further investigation was (and is!) warranted.
It is necessary to be cautious when comparing the scenarios of jet fuel burning and SAI. For one thing, a substantial portion of the jet fuel is burned at low altitudes in the vicinity of airports, leading to much higher concentrations of sulfates than were the distribution to be uniform throughout the atmosphere. So the next questions concerned, first, the relation between atmospheric concentration and toxicity; and second, the concentration of sulfates in the lower atmosphere resulting from SAI at scale.
Unfortunately, most of the focus of work on sulfuric acid toxicity [4] is on what is termed "strong acid mists", that is to say, high concentrations that might be encountered in industrial situations [5] or general studies of air pollution [6]. High concentrations are not relevant, or so one would hope. Eventually, however, the study by Pope, et al., Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution [7] , published in the Journal of the American Medical Association, was discovered, providing some answers. Quoting from its results, "Fine particulate and sulfur oxide–related pollution were associated with allcause, lung cancer, and cardiopulmonary mortality. Each 10-µg/m3 elevation in fine particulate air pollution was associated with approximately a 4%, 6%, and 8% increased risk of all-cause, cardiopulmonary, and lung cancer mortality, respectively." What then remained was to calculate the sea-level concentrations of fine particle sulfates resulting from annual injection of 10-20 megatons of sulfates into the stratosphere.
There are three obvious possibilities for performing such a calculation:
Calculating the concentration under the assumption that in the steady state resulting from sequential annual injections, an annual quantity of sulfates winds up uniformly distributed in both geography and altitude: a best case calculation in that it leads to the minimum concentration;
Using deposition estimates as a surrogate for concentration estimates.
Using a realistic global circulation model to calculate the eventual low altitude concentration, taking into account known atmospheric circulation and mixing phenomenology.
A calculation using the first of these approaches [8], assuming 10 megatons of SO2 injected at 20 km altitude, yielded a concentration of 1.47 micrograms per cubic meter, which is perhaps sufficient to induce a small but measurable increase in lung cancer mortality - not a good omen for SAI.
An indication of the variability of low altitude aerosol concentrations can be inferred from studies of deposition variability. Here, deposition refers to the amount of material deposited on the ground, per unit area, as a result of stratospheric injection. Unfortunately, the studies cited in earlier posts leading to more nearly optimal climatic responses did not include deposition estimates. Instead we use [9], which calculated deposition distributions for injection of 5 megatons of SO2 at tropical latitudes or 3 megatons in the arctic. One can easily calculate the deposition if uniform by simply dividing the amount injected by the area of the earth's surface, yielding 0.0147 grams SO4 per square meter per year for the 5 MT/year injection. The results of [9], converted to grams per square meter per year show a peak of 31.2 grams per square meter per year worst case, for an area in the Atlantic off the US east coast. Remembering that the injection quantity and location for [9] were atypical leads to tempering of conclusions as to the degree of concentration (here, by a factor of just over 2100) as well as the location of the spike. We can only surmise that localized adverse health effects may occur.
As I do not have access to a high fidelity global circulation model and lack the background in climatology to develop one on my own (much less have access to the kinds of computing resources necessary to run such a model), I must leave the third approach to others.
It may also be possible to estimate the adverse health effects of SAI using epidemiological data. According to [10] sulfur dioxide emissions peaked in 1979, at 136.6 million tons emitted, with the annual emissions decreasing on a global basis since then. In 2015 and 2022, the numbers were 81.1 and 69.3 million tons, respectively. Thus adding 10 - 20 million tons from SAI would result in a substantial reversal of the trend, and it might be possible to retrospectively observe the cost in terms of lives by correlating deaths with the increased emissions.
The current status of the issue is thus:
There is reason to believe that annual injection of 10 - 20 megatons of sulfates into the stratosphere could produce localized adverse health effects; and
There is also the possibility of such effects should another material be used.
The true degree of risk can only be determined through prediction of low altitude concentrations of SAI size-representative particles using high fidelity global circulation models that take results of SAI injection strategy investigations as input.
The next of my posts on SAI, on the topic of governance, will be the last. Or so it is hoped.
{Post updated 22:27 EDT on 8 May 24}
Notes
[1] The original image is at https://en.wikipedia.org/wiki/File:Small_cell_lung_cancer_-_cytology.jpg
[2] https://acp.copernicus.org/articles/16/10521/2016/acp-16-10521-2016.pdf
[3] https://fingfx.thomsonreuters.com/gfx/ce/xlbpgwkzkpq/GLOBAL%20AVIATION%20AND%20JET%20FUEL%20CONSUMPTION.pdf Incidentally, using 2018 data on jet fuel usage in conjunction with 1015 epidemiology is justified by the figures for mileage flown over the period, which was basically constant on an annual basis for the (large) markets surveyed.
[4] Sulfuric acid because the SO2 injected for SAI reacts with water vapor to form H2SO4.
[5] https://doi.org/10.15430/JCP.2019.24.3.139
[6] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7904962/pdf/nihms-1625966.pdf
[7] http://www.ssents.uvsq.fr/IMG/pdf/Pope_2002_1132.pdf
[8] As is my policy, details of the calculations will be provided to paid subscribers upon request.
[9] https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2009JD011918 as corrected by https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2010JD014579
[10] https://ourworldindata.org/grapher/so-emissions-by-world-region-in-million-tonnes