the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Mar 2025
Investigating the ability of satellite occultation instruments to monitor possible geoengineering experiments
Abstract. Solar radiation management is a method in the field of geoengineering that aims to modify the Earth's shortwave radiation budget. One idea is to inject sulphur dioxide or sulphuric acid into the stratosphere, where sulphate aerosols are then formed. Such experiments can probably be observed, for example, with satellite occultation instruments like SAGE III/ISS. The aim of the current study is to analyse, using MAECHAM5-HAM simulations and retrievals with the radiative transfer program SCIATRAN, whether it is possible to detect the formed stratospheric aerosols from emissions of 1 and 2 Tg S/y (sulphur per year) with the currently active satellite occultation instruments, taking into account an error estimate that is as realistic as possible. If these smaller amounts of sulphur are detectable, larger amounts will also be detectable. The calculations show that, considering the natural variability and the assumptions made here, the stratospheric aerosols formed from emissions of 1 and 2 Tg S/y in the quasi steady-state phase can be detected, which is not the case in the first month of the two-year initial phase.
- Preprint
(10945 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 25 Apr 2025)
-
CC1: 'Comment on egusphere-2025-1005', Travis N. Knepp, 14 Mar 2025
reply
Hello Anna! Thank you for submitting this paper.
The authors present a non-sophisticated (simplicity is beauty), simple evaluation of a solar occultation instrument's (here, SAGE III/ISS) ability to detect changes in stratospheric aerosol load with a continual injection of 1-2 Tg S/year. This study is elegant in its simplicity. Here, the authors determine that such changes would be detectable. To carry out this study the authors use MAECHAM5-HAM model to generate extinction coefficients at 500 and 550 nm, which were then converted to 520 nm (via Angstrom parameterisation), which was then fed into SCIATRAN to produce transmission data, which was then used to retrieve 520 nm extinction. I believe the accuracy of the authors' conclusions depends on the quality of these models, which I am not qualified to judge.
Overall, this paper is well written and I believe makes an important contribution to the scientific community. It would be interesting to see this study continue to evolve (e.g., if your "natural variability" included pyrocbs and volcanic eruptions (i.e., the stratospheric conditions that SAGE III/ISS has observed under), would you still have the requisite sensitivity to detect changes?; by continually injecting S you are changing the baseline stratospheric aerosol distribution...how does that impact particle growth and radiative transfer after another eruption, etc.), but the authors present an interesting and convincing proof of concept that stands on its own.
I only have minor suggestions for improving the paper.
1. In Fig. 1 (b), when does injection start? Did it start in January? I apologize if this was mentioned in the text and I missed it, but adding this information to the caption would aid the reader.
2. For Figures 2, 4, 5, 9, have the authors considered making the x-axis scale logarithmic? Especially in Fig. 4/5 this would help the reader appreciate the magnitude of change from background to enhanced conditions. Currently, I can barely tell that background is different from zero and a log scale would help the reader quantify this.
3. Do you have any information on how particle size may change with these injections? That level of information would be very interesting (at least to me).
Again, well done on the paper. I look forward to seeing this published; best of luck.
Citation: https://doi.org/10.5194/egusphere-2025-1005-CC1 -
RC1: 'Comment on egusphere-2025-1005', Anonymous Referee #1, 08 Apr 2025
reply
Review of Manuscript: " Investigating the ability of satellite occultation instruments to monitor possible geoengineering experiments"
The manuscript examines the ability of satellite occultation instruments to monitor potential geoengineering experiments using MAECHAM5-HAM simulations and retrievals with the SCIATRAN radiative transfer model. While the model experiments are straightforward and easy to interpret, I have two major concerns regarding the analyses presented in the paper.
First, the use of data from the initial two years of the simulation raises concerns about whether the model has reached an equilibrium state. It is crucial to ensure that the model's early-phase data are suitable for analysis. Could the authors provide insight into the model's performance during this period and demonstrate that it has stabilized?
Second, in line 214, the manuscript states, “The errors shown and discussed so far are based on one extinction profile per month.” It would be beneficial for the authors to clarify their rationale for this choice and to compare averaged profiles with those currently presented in the manuscript.
Beyond these major concerns, the manuscript is well-written and presents a novel approach by using a relatively small sulfur injection to assess the model's response. The study is valuable to the scientific community, and I have provided specific comments below that should be addressed before publication.
Comments:
- Lines 71-73: How is the model's dynamics treated? Is it prescribed using a climatology? Please elaborate.
- Lines 75-76: Could you clarify whether a single realization or an ensemble is used? If only one realization is run for 15 years, what is the spin-up time? If the first two years and the last three years are analyzed, is the model in equilibrium in the first two years (initial phase)? A time series showing equilibrium would be useful.
- Line 96: Correct the typo “trough” to “through.”
- Lines 97-98: Are these climatological values equivalent to those used in the model if the dynamics are prescribed?
- Lines 120-121: Is this statement based on observations or model experiments? Have you conducted a no-sulfur (0 Tg S/year) experiment to establish a background aerosol extinction coefficient? Please see my comment for Figure 7 (lines 187-188).
- Line 120: Insert "extinction" between "aerosol" and "coefficient" for clarity.
- Line 124: Could you please define “quasi-steady state” and “initial phase” more explicitly, especially in relation to model equilibrium.
- Lines 127-130: What is the reasoning behind showing these figures? Are they meant to highlight differences between the initial and quasi-steady-state phases? If so, please clarify.
- Lines 140-145: The sensitivity analysis is appreciated. Could you elaborate on why differences are larger below 15 km and improve above that altitude? Please do the same for the appendix figures.
- Lines 149-155: Why return to the initial-phase data here? If the quasi-steady-state phase is used for analysis, maintaining consistency would be preferable. I doubt if the initial phase data and their analyses are relevant as the model may not have attained equilibrium state by then (first two years).
- First line of section 3.2: Line 149 mentions that retrievals are based on initial-phase data, but this section focuses on quasi-steady-state data. Why mix the two?
- Third line of section 3.2: Could you explain why total errors for 1 Tg S/y are greater than for 2 Tg S/y in Figure 3? This seems counterintuitive.
- Line 162: Please clarify the term “true profiles.” Does this refer to model simulation results?
- Line 175: Please replace ± 550N with 55S-55N or something similar.
- Line 178: Do you have corresponding model simulation values for Figures 4 and 5? If so, please include them.
- Lines 180-181: Adding model-simulated SAOD using the 55S-55N latitude band would strengthen comparisons. Can you provide insight into the causes of maxima and minima in these figures?
- Lines 187-188: Section 3.2 is based on quasi-steady-state data, but here total errors seem to be derived from the initial phase. This is confusing—please clarify. Also, if you are analyzing a background case (0 Tg S/y) from the first month of the initial phase, a table in Section 2.1 listing all model simulations (0 Tg S/y, 1 Tg S/y, and 2 Tg S/y) would improve clarity.
- Lines 190-192: Could you please elaborate a bit more on as to why the total errors would be larger for Figure 7 compared to Figure 3?
- Lines 206-207: What is the rationale for using only the first month of the initial phase? How is that related to detectability of satellite occultation instruments? Would the model have reached equilibrium by then? Please see my comments above for line 75.
- Line 214: How was the selection made to use one extinction profile per month for Figures 2, 4, 5 and 9? It would be beneficial for the authors to clarify their rationale for this choice and to compare averaged profiles with those currently presented in the manuscript. Also, please consider showing model SAOD values in Figures.
- Lines 219-220: SAGE III/ISS measurements were not available before 2017. Please correct this.
- Lines 236-237: To highlight differences, please consider overlaying SAOD values from Figure 6 onto Figure 10 (e.g., plotting SAOD for 1 and 2 Tg cases at 50S and 30N from January to December).
- Lines 258-260: The discrepancy discussed here may result from the model’s non-equilibrium state during the initial phase. To ensure consistency, it would be preferable for the analyses to focus on data from the model’s equilibrium (quasi steady) state.
Citation: https://doi.org/10.5194/egusphere-2025-1005-RC1 -
RC2: 'Comment on egusphere-2025-1005', Anonymous Referee #2, 15 Apr 2025
reply
Review of the manuscript “Investigating the ability of satellite occultation instruments to monitor possible geoengineering experiments”, Lange et al.
Dear Editor, dear Authors,
The manuscript “Investigating the ability of satellite occultation instruments to monitor possible geoengineering experiments” by Lange et al. aims at analysing the expected capability of solar occultation instruments like SAGE III to detect and characterise stratospheric aerosol injection (SAI) geoengineering interventions. The topic of the manuscript is potentially important and relevant for a readership of satellite instrument and data scientists. Despite this, I have a few major concerns (especially Major Comment 1 below) about the manuscript and the inherent analyses that must be addressed before the manuscript can be considered for publication, in my opinion. I also have minor and specific comments, listed in the following. Please address the following major and minor comments and I will be willing to further evaluate a new manuscript version.
Regards.
Major Comments:
- The overarching motivation of this work is to assess the detectability of SAI interventions with SAGE III-like instruments. To do so, two geoengineering modelling experiments have been used as a pseudo-reality scenario, simulating SAI with injections of 1 or 2 Tg S/year. Now, different moderate stratospheric eruptions occurred in the last few years, e.g. Raikoke 2019, Hunga 2022, and more in the previous after-Pinatubo years, each with SO2 injection of that order of magnitude (1-2 Tg SO2 or less --> 0.5-1.0 Tg S or less). Many scientific works exist that show detection, tracking and quantitative characterisation of these plumes with SAGE III/ISS, e.g. More quantitatively from an aerosol extinction point of view, the perturbation in the aerosol extinction for this geoengineering experiment and study is of the order of 10-3 km-1 at 550 nm. This is quite comparable to perturbations of recent moderate eruptions like Raikoke, which sulphate aerosol perturbation was successfully detected studied and quantitatively characterised with SAGE III/ISS (e.g. Kloss et al., 2021, https://acp.copernicus.org/articles/21/535/2021/). So, even without this study, one could affirm that, yes, geoengineering plumes resulting from injection of 1 (or even less) Tg S and producing aerosol extinction perturbation of the order of 10-3 km-1 at 550 nm are observable and measurable with SAGE III. Thus, it is mandatory to clarify the scopes of this manuscript, restructure the manuscript accordingly and possibly change the title.
- The readership of interest for this kind of manuscript is more AMT that ACP, as it is very centred on instrumental aspects. Thus, I propose to transfer the manuscript to that journal.
- I don't really understand the background idea behind Sect. 3.3. Why is the error analysis performed for background case to study the initial phase of the perturbed case? Why not just making an error analysis of the initial phase itself instead? Why you conclude that "The background profile for 65° N (upper panel) is within the error range for all altitudes considered here, which leads to the conclusion that the emission of 1 Tg S/y cannot be observed at 65° N in January, i.e. the first month of the initial phase."? And why do you conclude that "the stratospheric aerosols formed in the first month of the initial phase can only be detected with satellite occultation instruments in a limited latitude range from – 10° to 14° N." All this is unclear to me and should be clarified.
- The quality of the text and clarity must be improved throughout the whole manuscript, more in the Specific Comments.
Minor Comments:
- L14: "and increase" --> "by increasing"
- L15: "SAI" is rather "stratospheric aerosol injection"
- L16 and Introduction in general: please consider citing more recent literature, e.g. https://www.frontiersin.org/journals/climate/articles/10.3389/fclim.2025.1507479/full
- L17: "sulphur" --> "sulphur dioxide"
- L19: "forcing" --> "radiative forcing"
- L20-21: "Simulations...W/m2" after how much time of 10 Tg S/y injections?
- L26-27: I don't get the logic here. Why now it is mentioned that it will be important to observe "small amounts of *sulphate aerosols" while before "large eruptions" where mentioned?
- L28 "SAGE" acronym not defined here. Please check if all acronyms are defined
- L29: "active" sounds odd here (SAGE is "passive" so this word can be taken in a different sense), also please mention "*solar occultation" and also why "one of the..."? Are there other solar occultation missions operational at the moment?
- L32; really there is nothing more recent than McCormick et al., 1979?
- L35: is 1-2 Tg S/y really a "small" injection? This sound like over a Raikoke-type eruption per year, so much larger than natural occurrence. In my opinion, this cannot be called "small". Please rephrase.
- -0.3 to -0.6 W/m2 is 10-20% the total greenhouse gases radiative forcing since preindustrial era, and up to twice the 10-years greenhouse gases radiative forcing; so why should this not be considered a significant radiative forcing?
- L39: again, these are not small amounts of SO2 injections but rather > 1 Raikoke per year...
- L55: "truncation at wavenumber 63" what does it mean?
- L66-67: What are the optical properties of sulphate aerosol in input to the radiation scheme? Are they calculated using an online Mie code and using the size distributions obtained with the model?
- L74: is sulphur injected as SO2? Which is the rate of SO2 injection - one shot per year or different injections finally cumulating to 1 or 2 Tg/y?
- Section 2.2: I don't think that SCIATRAN is properly introduced (what is it? how it works? relevant publications, etc)
- Eqs 1 and 2: can you mention here a typical obtained value for the Angström exponent?
- L104-105: "which is not publicly available": OK but is there at least a previous publication where it is used or this is the first one?
- L110: "The apriori state vector is kept constant" constant with respect to what? And constant at which value?
- 1: I'm not sure that the meaning of this figure is clear, especially because of the different color scales. Basically, why this figure is shown? To show how much the extinction coefficient is perturbed by SAI in the simulations? In this latter case, please use the same color scale for both panels
- L133-138: how are these modified settings chosen? It would be more significant to take variations that are comparable with uncertainties of O3, T, P from other (best available) measurements, and of realistic pointing errors.
- Eq. 6 is formally wrong because this is not expressed in %, as in the corresponding figure
- Fig.2: what does it mean, in the caption, the mention to "5°N"?
- I don't quite get the sense of sentence at L149-150
- Isn't it rather the square root of Eq. 7 what is actually called, before and after in the text, the "total error"?
- Section 3.2: can you quantify the typical total error in the stratosphere (the altitude of SAI injection)?
- L160: Remove "The following"
- L161: "January and July" of which year?
- Fig. 4: purple lines are difficult to see: why not another color, e.g. green?
- L164-166 and L180-185 (also Major Comment 1): the fact that these SAI interventions are visible with SAGE III is not really a surprise because these are not "small" injections but rather the same level of moderate eruptions, like Raikoke 2019, Hunga 2022, etc, which have been already successfully observed with SAGE III/ISS
- Section 3.3: see Major Comment 3
- L210: “Although…effect” see my previous comment on this
- L236-237: why not showing the geoengineering signal here (e.g. in Fig. 10)?
Citation: https://doi.org/10.5194/egusphere-2025-1005-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 153 | 28 | 5 | 186 | 5 | 3 |
- HTML: 153
- PDF: 28
- XML: 5
- Total: 186
- BibTeX: 5
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 118 | 47 |
| Germany | 2 | 51 | 20 |
| France | 3 | 21 | 8 |
| Canada | 4 | 9 | 3 |
| Belgium | 5 | 5 | 2 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 118