the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
In situ observations of gravity waves over tropical cyclone Hinnamnor (2022) suggest that TC gravity wave amplitudes are underestimated in ERA5 reanalysis
Abstract. Tropical cyclones (TCs) are known to regularly excite near-concentric outwardly radiating gravity waves in the lower stratosphere, with horizontal wavelengths of a few hundred kilometers. These disturbances can potentially drive in situ cirrus production, but their amplitudes have been poorly known: satellite observations produce substantially lower amplitudes than those in models or reanalysis, and there have been few aircraft overflights to provide in situ measurements. We present here data from a new TC overflight of supertyphoon Hinnamnor by the NASA WB-57 in 2022 during ACCLIP (the Asian Summer Monsoon Chemical and CLimate Impact Project). Hinnamnor produced a clear and large gravity wave temperature signal in the lower stratosphere, at 9.6 K peak-to-peak, nearly an order of magnitude larger than typical satellite studies and three times as large as in contemporaneous ERA5 reanalysis. The measured value would fall in the 97th percentile in ERA5, based on a sample of 114 summertime TCs in the Western Pacific basin, which produce near-ubiquitous but weaker concentric gravity waves. A metanalysis and comparison of ERA5 to observed winds suggests that both satellite retrievals and reanalysis underestimate TC gravity wave amplitudes because of resolution constraints. Satellite retrievals dampen short-vertical wavelength perturbations, while reanalyses fail to capture the rapid intensification of tropical cyclones that is associated with strong gravity wave production. These results imply that high-resolution measurements are needed to quantify the influence of TC-induced gravity waves on the stratospheric water budget.
- Preprint
(8769 KB) - Metadata XML
-
Supplement
(9817 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-4533', Anonymous Referee #1, 23 Sep 2026
-
RC2: 'Comment on egusphere-2026-4533', Anonymous Referee #2, 31 Oct 2026
Review of Kleinstern et al.
This manuscript presents unique and rare aircraft observations over a super-typhoon, Hinnamnor formed during the ACCLIP 2022 field campaign which was primarily conducted to sample the outflow of the Asian Summer Monsoon (ASM) using two aircrafts. In situ observations of temperature, ozone and water vapor were used to study the characteristics of typhoon-induced Concentric Gravity Waves (CGWs) along with Himawari-8 observations and ERA5 temperature data. The amplitude of CGW was derived from ERA5 data using two simple metrics developed by the authors and compared with that obtained from in situ data. They find that the CGW amplitude derived from in situ observations are three times larger than that derived from ERA5 reanalysis. Additionally, the authors compare their findings for typhoon Hinnamnor with 114 summertime typhoons formed in the Western Pacific basin using ERA5 database to give a broader context. They find that the amplitudes of CGWs generated by those typhoons were weaker in ERA5 data. Based on ERA5 winds and IBTrACS winds, TC intensity authors show that rapid intensification is underrepresented in ERA5. Authors emphasize the need to have high-resolution measurements for quantifying the influence of TC-induced GW on the stratospheric water budget.
Overall, the manuscript is well-written and the results are presented well. However, there are a few areas where the manuscript still needs improvements before it can be accepted for publication in ACP journal. Major strengths and weaknesses of this study are listed below before detailed comments and suggestions.
Strengths:
- The main strength of this study is having rare and unique in situ observations over a super-typhoon Hinnamnor. Such measurements are very essential to understand the dynamics and composition of the UTLS during the ASM and its interaction with typhoons whose timing coincide with the ASM.
- A total of 115 typhoons over the Western Pacific region are included in the analysis using ERA5 reanalysis to give a broader context to typhoon Hinnamnor. Authors have also developed two simple metrics for estimating the amplitudes of CGWs.
- Overall, the manuscript is well-written with detailed literature survey, good visualizations and discussion.
Weaknesses:
- The main weakness I found in this study is the under-utilization of available synergistic observations (in particular, GNSS-RO, ALADIN/Aeolus winds, and radiosondes data) with aircraft observation to comprehensively characterize the CGWs from Hinnamnor. Having such unique in situ data over a typhoon, both typhoon-induced dynamical and chemical impacts on the lower stratospheric composition could have been studied by combining satellite observations and model simulations making this study more aligned with the main objectives of the ACCLIP campaign. Additionally, other relevant observational studies are not well presented in this study.
- Several studies (Gong et al., 2015; Ratynski et al., 2025 and Wang et al., 2026) have already reported the presence of weaker CGWs in ECMWF reanalysis or ERA5 data with respect to observations. Additionally, prior studies have found that reanalyses do not accurately capture TC wind speeds (e.g. Schenkel and Hart, 2012) or intensification (Roberts et al., 2020; Baker et al., 2024) as mentioned by the authors in Summary section. Despite having the knowledge of these facts, this study appears to be more focussed on validating ERA5 reanalysis and providing support to previous studies rather than bringing new process-level understanding of typhoon-induced CGWs and their impacts on the UTLS composition.
- It is not clear if the two new methods developed by the authors are robust enough in estimating amplitudes of CGWs compared to the methods used in previous studies.
- The background winds in the lower stratosphere, despite being the most important drivers of CGWs (Wang et al., 2024) have not been included in the analysis to derive the characteristics of the CGWs.
Detailed comments and suggestions:
- Title: The current title indicates towards a case study but this study also includes analyses of 115 typhoons observed in ERA5 database. The title sounds like a concluding sentence rather than a heading. So, it needs to be modified.
- Lines 8-9: Which satellite-based studies are referred here, AIRS or GNSS-RO? Satellite-based (AIRS) studies usually sampled GWs in the upper stratosphere (between 30-40 km) as shown in Table 1 while aircraft measurements were confined in the lower stratosphere. Please justify the comparison done between the two measurements which differ a lot in their sampling volume, spatial and temporal resolutions.
- Lines 12-13: Did previous studies not report this limitation of ERA5? What makes this study different from the previous studies?
- Lines 25-26: Please add relevant references for typhoon induced structural changes in the UTLS (e.g., Biondi et al., 2013, 2015) and dehydration (e.g., Li et al., 2020).
- Line 54: Amplitudes of ~2-2.5 K were reported by Kim et al. (2009) in AIRS observations as listed in Table 1. So, the number “~1K” needs to be changed.
- Line 59: So, similar outcome is expected when reanalysis amplitudes are compared with in situ observations as found in this study. What new information does this study bring? Please cite those studies here. In Table 1, I see only one study by Kim et al. (2009) which provides the estimates of waves from both satellite and reanalysis at similar altitude region.
- Table 1 heading: Authors state that in most cases amplitudes are estimated from figures by eye. How accurate are those visual estimates of GW amplitude in comparison to peak-to-peak or other conventional methods?
- Please check: Dina or Hudah? Table 1 mentions the name of the typhoon as Dina whereas the citation says Hudah.
- There are several other studies (such as Gong et al., 2015; Kinoshita et al., 2026, etc.) which studied typhoon induced waves are not listed in Table 1. I suggest including those studies.
- Fig. 1: Why is the return flight track not shown here? Were there not any temperature measurements during the return flight? I am curious if the return flight captured any CGWs signature? Do Himawari-8 images show any signatures of CGWs?
- Lowest temperature was recorded at 06:54:34 UTC (~7 UTC) near the cloud edge at ~17.8 km whereas ERA5 temperature at 70 hPa (~18.8 km) at 6 UTC (~1 hour behind) is shown in Fig 1c which is nearly 1 km above the aircraft altitude. How does ERA5 temperature observations (137 levels) near 17.8 km look like?
- Where were the cold-point and dynamical tropopauses located?
- If the near-concentric rings were well beyond the typhoon cloud structure then why were not in situ observations between 0600-0630 UTC used for estimating GW amplitude?
- ERA5 hourly data are being compared with 20 Hz MMS temperature data that too at different levels, ~1 km apart in the UTLS. Is this a fair comparison? I suggest using high-resolution GNSS-RO or nearby radiosonde observations instead of ERA5 to have better comparison.
- Fig. S1: Please add a colourbar. Typo in the caption: replace “wihtout” with without
- Fig. 2: How does the ERA5 temperature vary along the flight track? It is important to have a comparison between ERA5 and MMS measurements. In addition to this, it is important to derive temperature anomaly along the flight track by using GNSS RO observations as done by Kim et al. (2016) to study gravity waves during the ATTREX field campaign.
- Moreover, the current analysis of GWs is restricted to 2D structures, I suggest using radiosonde observations available from nearby locations such as Guam, Naha, Okinawa or Osan during the passage of the typhoon to have better idea about the waves in vertical direction.
- Lines 82-86: NASA’s POSIDON field campaign in 2016 sampled the outflow of typhoon Haima (Jensen et al., 2020). However, typhoon induced gravity waves were not reported. I suggest exploring POSIDON data and extracting CGWs if possible. This would really strengthen the in-situ data base.
- Lines 89-92: If the data from those three campaigns are publicly available, its worth estimating GWs amplitudes. This would definitely populate in situ derived GW parameters in Table 1.
- Lines 92-94: Only one study based on radiosondes is listed in Table 1. What about other studies by Chane Ming? Authors mentioned 4 TCs in the text but cited only one (Dina) in Table 1.
- Kinoshita et al. (2026) also used radiosonde observations to study typhoon-induced gravity waves. Please look into this study and include it in Table 1.
- Lines 120-121: What about the return flight track?
- Lines 128-130: Were there any strong updrafts or downdrafts seen in MMS vertical wind velocities? Did measured horizontal winds show any signatures of GWs? How does ERA5 wind fields look like?
- Lines 131-133: Apart from ozone, information on cold-point and dynamical tropopause altitudes (which are missing here) are also important for stratospheric-tropospheric exchange processes induced by typhoons. Please consider adding them in the analysis.
- Lines 137-139: Was there any signature of overshooting convective hydration in isotopic composition of water vapor at flight level?
- Lines 153-154: ERA5 70 hPa is ~ 1 km higher than the flight level and it is expected to show large difference in temperature and hence the amplitude with respect to in situ data because ERA5 may not be able to capture small scale fluctuations in temperature within 1 km in the UTLS, especially during GW activity. So, this is not a fair comparison. I would rather suggest using nearest coincident GNSS-RO and radiosondes observations for better comparison.
- Lines 155-156: I did not find interpolated ERA5 data onto flight track for 31 August 2022 flight. Please help me find that plot.
- Lines 183-184: Why are traditional methods not applicable to ACCLIP aircraft data? Please explain.
- Gong et al., (2015) developed a ring detection algorithm to extract CGWs from AIRS radiance and ECMWF analysis temperature and provided global maps of CGWs. Have authors validated their methods with that method?
- Lines 205-212: Have these methods validated with other traditional methods? If so, please provide that information.
- Lines 219-221: Several peaks in ozone mixing ratio are seen during this half-hour stretch. Are these peaks wave structures? Have authors found similar features in ozone mixing ratios from satellite or ERA5 data? Are these fluctuations associated with STE processes or just measurement artifacts?
- Line 220: Please express 6.514 UTC in HH:MM format for consistency.
- Lines 222-225: Is 5.3 ppmv considered dry with respect to the background mean? What is the background mean water vapor mixing ratio at that altitude before and after the typhoon?
- How do ERA5 water vapor mixing ratio distribution look like at that level?
- Authors find no evidence of convective hydration or gravity wave dehydration. However, cirrus clouds are seen between ~16-17 km after the coldest point just before the dive (Fig. S2). Are these cirrus clouds not related to GW cooling/dehydration?
- What is the unit of total attenuated backscatter in Fig. S2?
- Regarding convective hydration, have authors looked at overshooting tops in Himawari-8 observations during the life cycle of the typhoon?
- Lines 227-231: Since vertical velocities from MMS are available. Adding a time series of vertical winds along the flight track in Fig. 2 will help corroborate these statements and will also help the readers to visualize the effects of GWs.
- Fig. 3a: Authors mention about comparison of mean ozone profile in Fig 3a. I could not find the mean ozone profile.
- Line 233: What do you mean by undilute gravity wave here?
- Line 234: Please express UTC time in HH:MM format for consistency.
- Lines 235-236: Please locate or label those three narrow bands in Fig. 3c. Are those outliers measurement artifacts? How can you rule out that possibility?
- Lines 237-239: Authors link ozone-temperature relationship to stratospheric gravity waves with strong vertical displacement without showing the measured vertical velocity along the flight track. Please show vertical velocity along the flight track.
- Several studies have suggested that typhoons bring clean marine ozone poor air to lower stratospheric levels (Li et al., 2020; Hanumanthu et al., 2020; Vogel et al., 2026 and references therein) during the Asian Summer Monsoon. This could be the case for the correlation between coldest point temperature and ozone mixing ratio values near 200 ppbv. Please provide more details on the possibility of such transport in the context of Hinnamnor by using satellite, reanalysis data or Lagrangian trajectories.
- Lines 244-246: What does this sentence (starting with “There is no..”) mean?
- Lines 249-253: Which methods were used to derive horizontal and vertical wavelengths? Are those visual estimates? Since this study uses both in situ and ERA5 data, its worth mentioning these estimates in Table 1. Authors have only included wave parameters derived from in situ data in Table 1.
- Fig. 4: Y-axis title will be “Temperature perturbation” instead of “Temperature”.
- Lines 257-258: Which figure is being referred here?
- Lines 262-264: There is no plot showing the variation of vertical velocities. Please show it to support this statement.
- Fig. 5: Why are the estimates of GW amplitude different in two methods? They differ by a factor of three. Why is this difference?
- Please mention the units for both x and y axes.
- Figure caption is too long. Please simplify it. Inferences from the figure can be included in the main text instead of the caption.
- Fig. 6: Wave amplitude derived from trimmed metric for in situ data shown by red dot is larger than what mentioned (2.54 K) in Table 2. Colourbar is missing in Fig. 6a, please add it.
- Section 3.4: A study by Wang et al. (2024) suggests that background winds in the lower stratosphere play most important role in affecting CGWs than TC intensity. I suggest analysing background stratospheric winds from aircraft, satellite (ALADIN/Aeolus) and ERA5 for Hinnamnor instead of using surface winds.
- Lines 345-347: What about observations from long duration balloons (Windborne Systems) and wind observations from satellites such as ALADIN/Aeolus?
- What are the recommendations for improving the representation of CGWs in ERA5 apart from having more high-resolution measurements?
- Lines 348-350: I suggest moving this sentence to the introduction and include this study in Table 1.
- Line 359: Change “e fforts” to “efforts”.
References:
Biondi, R., Ho, S.-P., Randel, W., Syndergaard, S., and Neubert, T.: Tropical cyclone cloud-top height and vertical temperature structure detection using GPS radio occultation measurements, J. Geophys. Res.-Atmos., 118, 5247–5259, https://doi.org/10.1002/jgrd.50448, 2013.
Biondi, R., Steiner, A. K., Kirchengast, G., and Rieckh, T.: Characterization of thermal structure and conditions for overshooting of tropical and extratropical cyclones with GPS radio occultation, Atmos. Chem. Phys., 15, 5181–5193, https://doi.org/10.5194/acp-15-5181-2015, 2015.
Gong, J., Yue, J., and Wu, D. L.: Global survey of concentric gravitywaves in AIRS images and ECMWF analysis, Journal of Geophysical Research: Atmospheres, 120, 2210–2228, doi:10.1002/2014JD022527, 2015.
Hanumanthu, S., Vogel, B., Müller, R., Brunamonti, S., Fadnavis, S., Li, D., Ölsner, P., Naja, M., Singh, B. B., Kumar, K. R., Sonbawne, S., Jauhiainen, H., Vömel, H., Luo, B., Jorge, T., Wienhold, F. G., Dirkson, R., and Peter, T.: Strong day-to-day variability of the Asian Tropopause Aerosol Layer (ATAL) in August 2016 at the Himalayan foothills, Atmos. Chem. Phys., 20, 14273–14302, https://doi.org/10.5194/acp-20-14273-2020, 2020. a, b
Jensen, E. J., Pan, L. L., Honomichl, S., Diskin, G. S., Krämer, M., Spelten, N., Günther, G., Hurst, D. F., Fujiwara, M., Vömel, H., Selkirk, H. B., Suzuki, J., Schwartz, M. J., and Smith, J. B.: Assessment of Observational Evidence for Direct Convective Hydration of the Lower Stratosphere, J. Geophys. Res.-Atmos., 125, e2020JD032793, https://doi.org/10.1029/2020JD032793, 2020.
Kim, J.-E., Alexander, M. J., Bui, T. P., Dean-Day, J. M., Lawson, R. P., Woods, S., Hlavka, D., Pfister, L., and Jensen, E. J.: Ubiquitous influence of waves on tropical high cirrus clouds, Geophys. Res. Lett., 43, 5895–5901, https://doi.org/10.1002/2016GL069293, 2016.
Kinoshita, T., Ogino, SY., Suzuki, J. et al. Study on a Typhoon-Induced Gravity Wave Captured by High-Altitude Radiosonde Observations in Okinawa in July 2024. SOLA 22, 10 (2026). https://doi.org/10.1007/s44393-026-00011-5
Li, D., Vogel, B., Müller, R., Bian, J., Günther, G., Ploeger, F., Li, Q., Zhang, J., Bai, Z., Vömel, H., and Riese, M.: Dehydration and low ozone in the tropopause layer over the Asian monsoon caused by tropical cyclones: Lagrangian transport calculations using ERA-Interim and ERA5 reanalysis data, Atmos. Chem. Phys., 20, 4133–4152, https://doi.org/10.5194/acp-20-4133-2020, 2020.
Ratynski, M., Khaykin, S., Hauchecorne, A., Alexander, J., Mariaccia, A., Keckhut, P., and Mangin, A.: Convection generated gravity waves in the tropical lower stratosphere from Aeolus wind profiling, GNSS-RO and ERA5 reanalysis,
https://doi.org/10.5194/egusphere-2025-394, 2025Vogel, B., Lauther, V., Köllner, F., Ekinci, F., Rolf, C., Strobel, J., van Luijt, R., Volk, C. M., Borrmann, S., Dragoneas, A., Eppers, O., Molleker, S., Hoor, P., Ort, L., Weyland, F., Zahn, A., Clemens, J., Günther, G., Kachula, O., Müller, R., Ploeger, F., and Riese, M.: Continental and marine source regions contributing to the outflow of the Asian summer monsoon anticyclone during the PHILEAS campaign in summer 2023, Atmos. Chem. Phys., 26, 6283–6319, https://doi.org/10.5194/acp-26-6283-2026, 2026.
Wang, X., Wang, Y., Zhang, L., Zhang, Y., and Guan, J.: A climatology of stratospheric gravity waves induced by tropical cyclones on the northwest Pacific Ocean, npj Climate and Atmospheric Science, 7, 1–11, https://doi.org/10.1038/s41612-024-00705-2, 2024.
Citation: https://doi.org/10.5194/egusphere-2026-4533-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 347 | 235 | 85 | 667 | 115 | 69 | 74 |
- HTML: 347
- PDF: 235
- XML: 85
- Total: 667
- Supplement: 115
- BibTeX: 69
- EndNote: 74
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript presents in situ aircraft observation of a large-amplitude gravity wave above TC Hinnamnor and shows that its temperature amplitude is substantially larger than in ERA5. The authors then place Hinnamnor in the context of a larger ERA5 TC sample and show that Hinnamnor is relatively typical in terms of gravity-wave activity. They further show that ERA5 underrepresents rapid TC intensification compared with IBTrACS, and suggest that this may contribute to the weaker gravity-wave amplitudes in ERA5 because rapid intensification is associated with enhanced gravity-wave production.
I appreciated the clarity of the manuscript and the figures.
I have a question about the method :
I was wondering if the authors could expend a bit on their definition of the 'background' used to isolate the gravity-wave signal. Subtracting a 2000–2022 daily climatology seems very low frequency to isolate GW. It removes the mean seasonal structure, but I am not sure if it would remove contemporaneous planetary-scale temperature anomalies? These larger-scale signals could still contribute to the variability measured. Could the authors comment on how sensitive the inferred GW amplitudes are to this background definition, or whether a more local or scale-selective detrending would give similar results?
I have a small rewording suggestion regarding the conclusion:
The abstract attributes the weaker gravity-wave amplitudes in reanalysis to resolution constraints, with the argument that ERA5 does not capture rapid TC intensification, which is associated with stronger gravity-wave production. I find this interpretation plausible and consistent with the previous literature cited in the manuscript, but the causal connection is a bit indirect, since near-surface wind speed is not itself the source of the gravity waves generated by TCs in the model. My understanding is that rapid intensification/stronger horizontal winds is being used here as a proxy for stronger convection, deeper heating, or stronger diabatic forcing, which would in turn generate stronger gravity waves? However, the current wording sometimes seems to link the horizontal wind evolution rather directly to gravity-wave production, without always making explicit the intermediate connection to the convective variables that actually control GW generation in ERA5. The abstract and discussion might benefit from slightly changed wording, to distinguish a plausible explanation from a demonstrated mechanism, but this is just a suggestion.