Review: https://doi.org/10.5194/acp-2021-81, A novel approach to sulfate geoengineering with surface emissions of carbonyl sulfide by Quaglia et al.

Overview:  This paper proposes using carbonyl sulfide (COS) released at the surface for increasing stratospheric sulfate aerosol.  This proposal has the advantage of not requiring distribution by stratospheric aircraft.  The paper gives a great amount of detail on the differences between the three scenarios examined, a background, a geoengineering scenario using SO2 and one using COS.   It is generally well written.  I recommend publication after addressing the comments below, and most importantly, comment #1.

1)   Crutzen (2006) https://link.springer.com/article/10.1007/s10584-006-9101-y  states “An alternative may be to release a S-containing gas at the earth’s surface, or better from balloons, in the tropical stratosphere. A gas one might think of is COS, which may be the main source of the stratospheric sulfate layer during low activity volcanic periods (Crutzen, 1976), although this is debated (Chin and Davis, 1993). However, about 75% of the COS emitted will be taken up by plants, with unknown long-term ecological consequences, 22% is removed by reaction with OH, mostly in the troposphere, and only 5% reaches the stratosphere to produce SO2 and sulfate particles (Chin and Davis, 1993). Consequently, releasing COS at the ground is not recommended.” 

Based on this, first,  this paper should note that this has been proposed before (and therefore may not be so novel) , and discounted due to potential ecological damage.  There is discussion at the end of the paper that this sort of thing should be looked at.  That discussion should be at the start of the paper.  Second, is the balance between global warming due to COS vs cooling due to aerosols produced taken into account in this study?  Bruhl et al. 2012 state “Further, using a chemical radiative convective model and recent spectra, we compute that the direct radiative forcing efficiency by 1 kg of COS is 724 times that of 1 kg CO2. Considering an anthropogenic fraction of 30 % (derived from ice core data), this translates into an overall direct radiative forcing by COS of 0.003 W m−2 . The direct global warming potentials of COS over time horizons of 20 and 100 yr are GWP(20 yr) = 97 and GWP(100 yr) = 27, respectively (by mass). Furthermore, stratospheric aerosol particles produced by the photolysis of COS (chemical feedback) contribute to a negative direct solar radiative forcing, which in the CCM amounts to −0.007 W m−2 at the top of the atmosphere for the anthropogenic fraction, more than two times the direct warming forcing of COS. Considering that the lifetime of COS is twice that of stratospheric aerosols the warming and cooling tendencies approximately cancel.”  They also say “Therefore, if we account for indirect chemical effects in GWP calculations, also customary for gases such as methane (IPCC, 2007), it follows that COS has almost no net climate impact.”  So, if the warming and cooling tendencies cancel, is there an actual advantage to increasing COS emissions?    At line 280 this paper states that the COS case produces an RF of .17 W/m2.  What is the RF due to the increased sulfate aerosol produced by the COS?  How do these results compare to the conclusions of Bruhl et al. 2012?

2) The text makes the statement “it is not a toxic gas for humans or ecosystems: negative effects have been found only when concentration exceed 50 ppm (100,000 times more than the background mixing ratio (Kilburn and Warshaw, 1995; Bartholomaeus and Haritos, 2006).”  The Bartholomaeus and Haritos paper does not have that statement, and the Kilburn and Warshaw paper is about H2S and not COS.  Different references are needed to say that there are no negative effects on plants and ecosystems.

3) The Budyko, 1978 reference link (Budyko, M. I.: The Climate of the Future, American Geophysical Union, https://doi.org/10.1002/9781118665251.ch7, 1978.) is broken (it says the doi cannot be found, this is apparently an error on the AGU web page.).  I believe the publication date is actually 1977, and you can get to the full book at https://agupubs.onlinelibrary.wiley.com/doi/book/10.1029/SP010.  The reference should be to the full book, since chapter 7 doesn’t address the point being made. And, in support of the text (actually proposing SRM) NAS 1992 (National Academy of Sciences (NAS): 1992, Policy Implications of Greenhouse Warming: Mitigation, Adaptation, and the Science Base, Panel on Policy Implications of Greenhouse Warming, Committee on Science, Engineering, and Public Policy, National Academy Press, Washington DC, 918 pp, https://www.nap.edu/catalog/1605/policy-implications-of-greenhouse-warming-mitigation-adaptation-and-the-science should be added.

4) Line 72:  It should be noted at the beginning of the paper how the increase of 40 Tg-S/yr compares to current emissions.  I finally found the factor of 400 in the conclusions.

5) Line 105-106 says “This means an increase of 0.8 ppbv with respect to background condition, that would produce a direct RF negligible if compared to other well mixed greenhouse gases.”  First, change to “a direct RF that is negligible compared to other well mixed greenhouse gases.”  Second, is the RF negligible to the negative forcing caused by the increased stratospheric aerosols?

6) And, another experiment that could be run.  Instead of emitting a large amount of OCS at the surface, what about emitting at the tropical tropopause?   It would require significantly less material, and may produce similar results, with a higher aerosol layer and similar latitudinal distribution.  It would also avoid any issues with ecological damage due to increasing surface COS amounts.

This study discusses a new potential approach for sulfate geoengineering using an enhancement of COS at the surface. The study is very well organized, clearly written, and nicely presented. The paper should be published as an important contribution to possible approaches for sulfate geoengineering and their effects. I am supporting the publication after the authors considered the following points listed below. While this is a minor revision in terms of workload, I am suggesting major additions to the discussion and the abstract.

Here are some major concerns considering this new approach that has not been clearly addressed in the paper:

1. COS is more uniformly distributed in the troposphere and has a much longer lifetime than sulfate. This however suggests that there is much less control using COS than for example using SO₂ injections for sulfate geoengineering and limits the potential use of a feedback control algorithm to modulate the amount of cooling in different hemispheres. However, earlier studies by Kravitz et al. (2017) have shown the (game-changing) potential of using a feedback control algorithm to reduce side-effects for instance to reach surface temperature targets or other impact-relevant targets. This is a major drawback of this approach in addition to the lack of the rapid regulation of the injections in case of a large volcanic eruption as already pointed out in this paper.
2. Due to its toxicity, there is seems to be a hard limit in using COS. This study increased surface COS concentrations to reach 35ppb, which is very close to the possible limits of 50ppb, as stated in the text. The first concern that needs to be at least mentioned for future work is to identify how reliable the studies are that estimate this toxicity considering potential long-term increases of COS. Secondly, the enhancement of COS reduces AOD by 0.08 with a radiative forcing of -1.3W/m2, which may be translated to less than 0.5 degrees of cooling. Due to the uncertainty in different models, the cooling could be less and may not be sufficient. Furthermore, due to the long phase-in time, it will take a long time to find out how much cooling can be achieved. What if more cooling is needed? Would COS be eventually replaced by SO2 injections? What is the point then of using it in the first place, while adding the danger of exposing humans and ecosystems to a toxic pollutant? To me, this is a major issue that may be a showstopper for considering this method.
3. It needs to be more clearly stated that surface UV is largely reduced with this method, and this can be harmful to humans and ecosystems. A more detailed discussion with the region and season would be helpful.

Detailed comments:

Abstract: The last sentence in the abstract stated that COS emissions are feasible. It may be technically feasible, but I think, the authors need to also point to the drawbacks of this approach in the abstract and conclusions, including the limitations compared to stratospheric SO2 injections.

Line 19: I don’t think, this type of intervention can be classified as a “short-term” intervention, earlier studies have shown, that even a “Peakshaving” scenario may require injections between 80-160 years (Tilmes et al., 2016, 2020).

Line 22: “optically active” is somewhat strange. You could maybe say, “the aerosol layer is thickened and therefore reflects more sunlight…“

Line 27: I don’t understand what you mean with “any proposed compound would quickly react to form sulfate aerosols”, only sulfur will form sulfate aerosols. Other components may be coded with sulfates that are in the atmosphere. Is that what you mean?

Line 38: Is this the tropospheric lifetime? What is the stratospheric lifetime.

Line 65: Since you are looking at UV, please also state what photolysis scheme is used in the model and if photolysis varies with aerosol concentrations.

Line 72: Is there a reason why increase COS emissions are placed at the same locations as the anthropogenic emission?

Line 101: Could you add an estimate of how much surface cooling one expects with an increase in 0.08 AOD? I would be probably less than half a degree of cooling considering the related GeoMIP experiments, is that correct?

Line 104: Didn’t you state that the lifetime is 35 years, not 3.8 years? What is different here?

Line 106-107. I don’t follow, what is meant here. 0.8ppbv of what, and why do you refer to the RF of other greenhouse gases? Do you mean, that after 20years, COS values have declined close enough to the background to not have a significant impact on the RF?

Line 230: The catalytic NOx cycle is decreased with more surface area density, which results in less ozone loss. Isn’t the inhibition of denitrification more important in high latitudes and for cold temperatures, and less in other latitudes

Line 234: “its photodissociation” what is “its” referred to here?

Line 240: change “UVA decrease is everywhere negative in both SG experiments” to “UVA decreases everywhere in both SG experiments”

Table5 and Figure 8: You are nicely showing total column ozone variations with region and seasons, and it is very clear that there are differences in sign in the response. Illustrating UV changes annually is not a very meaningful measure especially for high latitudes. I would strongly recommend expanding this figure and showing 4 seasons instead or in addition to the annual values. I would also expand the discussion on the UV impacts, and why UV-A and UV-B and shown separately. Not sure if there is any reference to Figure 8 in the text?

Line 269: reference should be Brühl et al. (2012)

Line 289:  should be (Figures 9 and S6)