| 21 Jul 2026
Observed Long-Term Changes in Global Tropopause Characteristics
Abstract. Variability or long-term change in tropopause characteristics has important implications for climate, especially since its temperature – to a large extent – dictates the amount of water vapor (a critical greenhouse gas) in the stratosphere. While conventional observations sufficient for examining tropopause characteristics provide limited geographic coverage, the growing global navigation satellite system radio occultation record provides relatively unlimited geographic coverage continuously since early 2001. This study applies the temperature lapse-rate tropopause (LRT) and potential temperature gradient tropopause (PTGT) definitions to the most recent 23-year radio occultation data period (2003–2025, inclusive) to create robust climatologies and assess long-term changes (statistically significant positive/negative trends) in the altitude, temperature, and sharpness of the primary tropopause and in the occurrence of multiple tropopauses. Regardless of definition, primary tropopause altitudes and temperatures are found to be increasing in most locations globally, consistent with other studies. A narrow region of decreasing tropopause altitudes in the subtropics and over the southern hemisphere east Pacific ocean basin is found to be driven by localized tropical narrowing, while the width of the tropics (based on tropopause altitude) is found to be increasing otherwise, especially throughout the northern hemisphere. Primary tropopause sharpness is also found to be increasing in most locations, but especially so in the tropics, where it is driven by both upper troposphere stability decreases and lower stratosphere stability increases. Finally, double and triple tropopauses are becoming more frequent in both the tropics and midlatitudes.
Review of „Observed Long-Term Changes in Global Tropopause Characteristics“ by Homeyer and Tinney
This study investigates present trends of tropopause characteristics from GNSS radio occultation measurements for the time period 2003-2025, including height and temperature of the primary tropopause, tropical width, frequency of double and triple tropopauses and tropopause sharpness. Specifically, the authors focus on regional differences and claim to be the first to address tropopause sharpness trends as well as multiple tropopause trends from GNSS-RO observations. The mentioned tropopause characteristics are compared for two (three) tropopause definitions in the extratropics (tropics), namely the commonly used WMO lapse rate tropopause (LRT), the more recent potential temperature gradient tropopause (PTGT) and the cold point tropopause (CPT) in the tropics. The in-depth comparison of LRT and PTGT features is another novelty of the article. As such, the paper makes an important contribution to our understanding of the tropopause and its undergoing changes.
The manuscript is already in good shape. I suggest only minor revisions. While the conclusions already offer some discussion and outlook, certain statements throughout the manuscript would benefit from further elaboration, and the placement of the findings within the broader literature could be expanded in a few places (see detailed comments below).
My major criticism concerns the generalization that the findings are (apparently) insensitive to the tropopause definition (e.g., L8 and L274-275). While this holds for the definitions considered here, it should be made clear that the same behavior cannot necessarily be expected for dynamical or chemical tropopaus definitions. Furthermore, given the close similarity of the results obtained with the LRT and PTGT definitions, the advantages of PTGT over LRT the advantages of PTGT over LRT should be articulated more explicitly (rather than only referring to Tinney et al. (2022)).
In addition, I recommend adding a code availability statement.
Details:
L8: In my opinion, it should be made clear that „regardless of the definition“ means: LRT, PTGT (and CPT in the tropics).
L22: For completeness, a reference regarding STE should be added.
L63-68: Beyond tropopause-break-based proxies, other approaches to characterizing tropical width and its changes exist (e.g. Davis and Birner, 2017; Waugh, 2018; Turhal et al., 2024). Notably, Turhal et al. (2024) and Weyland et al. (2025) report evidence of tropical narrowing in the SH based on zonal mean data and even a shape change of the tropopause break.
L78: What about studies based on reanalyses? Trends derived from reanalysis data must, of course, be interpreted with caution. However, since key variables such as temperature are constrained by the assimilation of (satellite) observations, tropopause characteristics are generally captured fairly well in modern reanalysis products.
L87: It would be helpful for the reader to make it clear at this point that tropical tropopause frequency is considered here to examine tropopause break latitude. You could simply move the brackets from line 92 up.
L119-120: In my opinion, these failure cases should be specified here (instead of merely referring to Tinney et al. (2022)), not least to highlight the advantages of the PTGT. How are these failure cases handled/why can the PTGT resolve these failure cases?
L169: A reference should be included here.
L179-181: The CPT altitude trends are (significantly) positive at the locations where LRT and PTGT trends are negative. This should be mentioned and discussed briefly.
L185: Personally, I would prefer a title that clearly states that this section addresses the tropopause break latitudes/tropical width trends.
L198: I suggest showing the mentioned global frequency distribution. For example together with Figure 7.
L210-212: A brief discussion of possible causes for this local tropical narrowing would be interesting.
Figure 6: The color bar in panel a) is not ideal. The CPT altitude appears to be very uniform.
L216: A reference should be added here.
L220-222: It would be interesting to have a short discussion on why these local maxima exist.
L225: „most locations equatorward of 50°“, I would say.
L229-230: This is an interesting observation and hypothesis, worthy of further discussion.
Could you elaborate on how the changing tropopause sharpness and the jet structure are related and add the respective references?
Figure 9: In this figure as well, the color bar limit is not ideal. Consider choosing a slightly higher limit.
L241-243: Here as well, it would be worthwhile to discuss hypotheses for the observed sharpness changes (already before the conclusion).
L247: Regarding the wording: In my opinion, the motivation to address (or not address) multiple tropopauses shouldn’t be that it has (not) been covered in previous studies, but rather if its generally of physical interest/of any added value compared to double and triple tropopauses.
Figures 13 and 14: Consider merging the two figures in one for easier comparison.
L248-252: I find it interesting that even though the climatological tropical width is different between PTGT and LRT (Fig. 2), the latitudes of the highest double tropopause frequencies are basically identical for PTGT and LRT (Fig 13). Maybe you could discuss this briefly?
L258-260: Is one of the definitions closer to reality in your opinion? Maybe you could elaborate.
L268-269: This statement is very vague. Could you specify the mechanisms?
L273: ...and the CPT in the tropics.
L274-275: Again, the observations might be insensitive to the tropopause definitions considered in this study, but not necessarily any tropopause definition in general.
L286-288: From Fig. 10 it’s not evident to me that the sharpness trends are confined to the tropics and mid-latitudes. Trends in the subtropics look similar. This is also stated in L303-304.
L301-302: I agree that it can be beneficial to consider tropopause trends relative to the tropopause break, depending on the research question. Locally, however, the average tropopause altitude does indeed increase with increasing tropical tropopause frequency.
References:
Davis and Birner, 2017: https://doi.org/10.1175/JCLI-D-16-0371.1
Turhal et al., 2024: https://doi.org/10.5194/acp-24-13653-2024
Waugh et al., 2018: https://doi.org/10.1175/JCLI-D-18-0108.1
Weyland et al., 2025: https://doi.org/10.5194/acp-25-1227-2025