| 18 Jun 2026
Dependence of Atmospheric and Climate Impacts on Launch Latitude and Seasonal Variation in Rocket Emissions
Abstract. Rocket launch emissions, including black carbon (BC), water vapor, and nitrogen oxides, can perturb atmospheric composition and climate, with impacts expected to increase under future growth in launch activity. Previous studies have shown that rocket-emitted BC can warm the stratosphere and alter ozone (O3), but the dependence of these impacts on launch latitude and seasonal timing remains poorly understood. We investigate how launch latitude and seasonality influence the steady-state distribution of BC and the resulting climate responses. We use the Whole Atmosphere Community Climate Model version 6 to simulate emissions from a medium-lift kerosene launch vehicle at six launch latitudes: 55° S, 29° S, 0° N, 29° N, 55° N, and 70° N. Year-round emissions are applied at all latitudes, while boreal summer-only and boreal winter-only scenarios are applied at selected latitudes using the same annual emission rate, corresponding to 30 Gg yr-1 of BC over approximately 11–70 km. The results show that Southern Hemisphere (SH) launches produce larger global BC burdens and stronger stratospheric warming, whereas O3 depletion is stronger for Northern Hemisphere launches. Launch seasonality modifies BC transport. Low-latitude boreal summer launches transport more BC into the SH, leading to greater stratospheric warming but less O3 depletion. At higher launch latitudes, launch season controls the altitude of the steady-state BC. These findings indicate that the climate impacts of launch activity depend not only on emission magnitude but also on the latitude and timing of launches.
The manuscript presents a model study of how rocket launches impact the stratosphere, and presents asymmetries in the results per hemisphere, as well as highlights the different impacts of year-round vs. seasonal launches. The presented results are interesting and worth publication, but some of the figures are not very informative due to lack of appropriate statistics, as explained below. Further, the key outcome of the hemispheric asymmetry is presented, but some deeper analysis on the mechanisms that cause it would be useful. I recommend major revisions before considering acceptance.
General comments:
The longitude dependence on rocket launches climate impacts is completely ignored in the manuscript. Although indeed latitude might matter more, when looking at seasonal and non-continuous emissions the longitude might have a non-negligible signal on where the effects happen. This could be particularly relevant in the troposphere (which the authors largely do not affect, since emissions start at 11 km), but also in the stratosphere (e.g. QBO). A couple statements about this would be useful.
The ozone changes presented as a function of location and season of launches are interesting. A thing that is not discussed much, but it is strongly complementary to the temperature changes, is the availability of light. Boreal winter-only launches will have less light available to promote photochemistry in the northern hemisphere, and much more in the southern one. I believe these changes should be discussed alongside the temperature changes. This is also important for shortwave radiative forcing; BC will have a very small impact in the dark. Also, what is the role of heterogeneous chemistry on BC surfaces that promotes ozone loss? This is not discussed at all.
Figures 1, 3, and 4, are missing error bars in both axes; they are not fully informative the way they stand right now, and the polynomial fits of figures 3 and 4 are more misleading than useful, since they imply trend shapes that might not be real, especially in figure 4. The weighted mean latitude is particularly misleading to show as one dot without a range, when BC is distributed over a very broad area. Since these figures support most of the conclusions here, the results in them need to be carefully and accurately presented.
Specific comments:
Line 34: Although the statement about NOx and ozone depletion is correct, it is important to mention that regarding rocket launches the NOx emissions are minimal in the stratosphere where most of the ozone is.
Lines 55-57: Although interplanetary missions have indeed very narrow launch windows, 1) these windows have no fixed seasonal cycle when looking at different years, since the planets’ orbits are not in sync with that of Earth, so over multiple years they won’t create a seasonal bias, and b) very few launches are actually interplanetary; the vast majority of them are to low Earth orbit.
Line 65: I believe it is incorrect to state that the Australian wildfires that emitted BC in the lower stratosphere are comparable with rocket launches that emit BC up to 70 km (in this study) or higher. The comparable magnitude (order of magnitude; 22.5 and 30 are 33% different) is coincidental, and the injected profiles are different enough, both vertically but also temporally (one event vs. multiple launches) to not be comparable at all.
Lines 66-71: 2000 launches by 2030 result in 4.25 Gg/yr BC, which is 1/7th of the emissions in the manuscript, so are you assuming 14000 launches for near-present-day? This sounds like orders of magnitude off. Later (line 119) 10000 launches are mentioned, which is slightly inconsistent with the calculation above, can you please clarify? Since the number of launches has been raised earlier from other studies, please state here now many launches you are assuming per year.
Line 94: 0.134 micrometers is the size at emission time? Because MAM4 should change that in the mixed particles, right?
Line 128: There are 4 seasons, so 10 Gg/season would be equivalent to 40 Gg/yr, which is inconsistent. Even more, the motivation of the different seasonality of launches is to evaluate what would happen if the space industry avoids certain seasons for whatever reason, which is a very unlikely scenario in my opinion; using less emissions though means less launches, which is likely unrealistic, if the space industry wants to compress their launches to a season only but keep the number of launches fixed per year, to maintain their space fleet in business-as-usual conditions.
Section 2.3: It would be useful to show a typical emissions profile here as a figure, similar to what Maloney et al. (2022) has presented. Also, how realistic is the assumption that the rocket goes to geostationary orbit (GEO), when the vast majority of launches go to low Earth orbit (LEO)? How different would the emissions profile be for an ascent to LEO instead of GEO? Lastly, are you emitting at every time step during the emissions period (year or season) and from all sites at once? So the number of simulations is 4 (control, year-round, boreal summer, boreal winter)? Plus the extra one in Table 1. This is not clearly mentioned, and the way Table 1 is structured, it might imply one simulation per tick mark.
Line 163: Where does the value 25 g/kg come from?
Lines 183-185: why the inconsistency in the upper altitude bound between stratospheric responses and burden calculations?
Beginning of section 3 and figures 1a-c: How do you calculate the mean stratospheric temperature, and what is the meaning of it in the first place? There is a huge latitudinal and vertical gradient, so what is the point of calculating a mean of that? As an example, a temperature change inside the ozone layer would have a very different impact from a temperature response above it, in the upper stratosphere. Importantly, how do you calculate the average in the vertical, what kind of weighting do you use?
Figure 2 shows that 5 years are not enough to reach a proper steady state. I would recommend using 10 years as a spinup, and analyze the last 10 years, or even extend the simulations a bit to include more equilibrated years, if 15 (or more) is what you prefer. The deviation from the equilibrium between years 5-10 might seem small, but it will contribute to a variability range that is larger than what it should be.
Also in figure 2, I am a little surprised by how little BC is spreading, e.g. in 2d and 2f. Can you please comment? Are the particles too big and they sediment quickly, or there is another explanation? If 2f particularly, why the transport does not happen symmetrically across the two hemispheres?
I do not understand the results in figure 3b. The launch latitudes span from 60S to 60N, but the weighted mean latitude of BC goes only from 40S to 40N. This sounds backwards from what Brewer-Dobson circulation (BDC) would do. Can you please explain? For figure 3a, what is the mechanism that favors BC load to be higher in the Southern hemisphere? BDC is quite symmetric, so why BC appears to have a longer lifetime in the south? Your comment in lines 211-212 implies asymmetric hemispheres, please demonstrate that, since it is the main conclusion of this work, present even in the title.
Lines 228-229: You could mention here that this is also found by Tsigaridis et al. 2024, already cited elsewhere in the manuscript.
Technical corrections:
Table 1: you might consider reversing the order of rows, to be more intuitive geographically (Northern Hemisphere is up). Also, the explanation of the star should appear as a footnote of the table, not in the caption.