the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Feb 2025
The dilemma in identifying WMO-defined tropopause height using high-resolution radiosondes
Abstract. The tropopause plays a critical role in stratosphere–troposphere exchange and climate change. Its height is conventionally defined based on the World Meteorological Organization (WMO) threshold temperature gradient, yet this gradient is intrinsically linked to vertical resolution. Data with higher vertical resolution inevitably reveal finer gradient structures. While in situ radiosonde temperature measurements are considered the most reliable source for tropopause structure, high-resolution (5–10 m) soundings would be expected to yield more precise height estimates. The near-global coverage of high-resolution radiosondes, accumulated over even decades, promises valuable insights into long-term tropopause variability. However, our analysis demonstrates that the original WMO definition can lead to an underestimation of the tropopause height when using high-resolution soundings, potentially misidentifying the tropopause within thin inversions or temperature gradient discontinuities below tropopause. To address this, we leverage ERA5 tropopause heights as a reference to develop a high-resolution-optimized method. We evaluate three methods: original WMO method, Moving average method, and Coarse–Fine method. The results reveal that the mean differences between the three methods and ERA5 were 800 m, 280 m, and 180 m, respectively. Notably, ERA5 systematically overestimated the tropopause height compared to all methods, with this discrepancy particularly pronounced in the edges of the Hadley circulation. The proposed Coarse–Fine method, by effectively bypassing thin inversions and gradient extrema while preserving the fine–scale structure of the tropopause height, presents a promising tool for future investigations into long-term tropopause trends.
- Preprint
(7914 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 24 Mar 2025)
-
RC1: 'Comment on egusphere-2024-4198', Anonymous Referee #1, 28 Feb 2025
reply
Review of 'The dilemma in identifying WMO-defined tropopause height using high-resolution radiosondes' by Yu Gou et al.
General comments
This new study by Gou et al. explores the challenges in identifying the tropopause height using high-resolution radiosonde data based on the World Meteorological Organization (WMO) definition. The study highlights that the original WMO approach tends to underestimate the tropopause height in high-resolution datasets due to the presence of thin temperature inversions and gradient discontinuities. To address this issue, the authors propose two alternative methods, the Moving Average (MV) method and the Coarse-Fine (C-F) method, both of which provide more consistent tropopause height estimates compared to ERA5 reanalysis data. It is found that ERA5 systematically overestimates the tropopause height, particularly near the Hadley circulation edges, while the WMO-defined method underestimates it. The C-F method emerges as the most effective in preserving fine-scale structures while filtering out spurious lower-altitude tropopauses. The study concludes that modifications to the WMO method are necessary when using high-resolution radiosonde data for accurate long-term tropopause trend analysis.
The study addresses a crucial problem with practical implications for climate research and meteorology. The research employs robust statistical techniques, sensitivity analyses, and cross-validation with ERA5, ensuring a high level of scientific rigor. The methodology is well-described, but a deeper discussion on the physical basis of biases and additional validation with independent datasets could enhance transparency. The manuscript is mostly clear and concise. I would like to recommend that the paper be considered for publication, subject to the minor comments listed below.
Specific comments
l40: It is a little unclear to me what you mean by 'constant emitted temperature'?
l106: The dynamical tropopause in the tropics is usually defined by a potential temperature threshold, not by potential vorticity. Please clarify.
l116: Lapse rate is defined as temperature difference over height difference. However, for radiosondes this is probably calculated via pressure differences? Could you please clarify and elaborate?
l119: Reference paper for ERA5 should be cited: Hersbach H, Bell B, Berrisford P, et al. The ERA5 global reanalysis. Q J R Meteorol Soc. 2020; 146: 1999-2049. https://doi.org/10.1002/qj.3803
l136: Please clarify the shape of the moving window. Presumably a boxcar/tophat function was used?
l195: It would be good to know if the high resolution radiosonde data (or a downsampled version of the data) have been assimilated into ERA5? Presumably the data are not independent?
Fig. 6: These maps suggest that the statistical differences are due to a few individual outliers, probably related to inversions, rather than a bias across all profiles.
Technical corrections
l17: below _the_ tropopause
l19: Movering _A_verage method (here and in other places in the manuscript)
l63: 'Instead' -> 'In contrast'
l69: high-resolved radiosonde _data_ can
l69: understanding of __ tropospheric processes
l73: replace 'Hoffmann et al. (2022)' by 'Hoffmann and Spang (2022)' throughout the manuscript
l82: 'directly WMO-defined method' -> 'original WMO method to detect the tropopause' (the term 'WMO direct method', WDM, is defined only later in the manuscript)
l92: regularly __ twice a day
l97: avoid repetition of 'typically'
l132: occurr _at_ higher latitude
l154: _Sensitivity_ analysis
l160 and l172: replace 'enhancement' by 'increase'
l206: revisit/rephrase 'in the austral with equivalent latitude band'
l224: suggest section title 'Discussion and conclusions' (since discussion comes first)
l241: approximately 300 m _in the upper troposphere and lower stratosphere_ (for clarity)
l265: fail to _avoid_ the
Fig. 2: y-axis label 'Hight' -> 'Height'
Fig. 5: Scatter plots would be better visible if they would be shown next to the maps, and not on top of them. Replace 'in the lower side' by 'at the top of the scatter plots' in figure caption.
Citation: https://doi.org/10.5194/egusphere-2024-4198-RC1 -
RC2: 'Comment on egusphere-2024-4198', Anonymous Referee #2, 14 Mar 2025
reply
The authors leverage a recent year of high-resolution (5-10 m) radiosonde observations and ERA5 reanalysis output to explore application of the conventional World Meteorological Organization (WMO) lapse-rate tropopause definition to these datasets. The authors argue that the WMO definition fails to reliably identify tropopause height using high-resolution data and develop several alternative means of application to improve performance. Finally, they compare their results to those based on ERA5 reanalysis and report upon outcomes. While the study is generally well-structured, the analysis unfortunately is built upon an incorrect implementation of the WMO definition which undermines the fidelity of all analyses presented. Thus, I cannot recommend publication of this study, but encourage the authors to repeat the analysis with correct implementation of the WMO definition (and others) and continue to pursue the comparisons with ERA5 reanalysis, which are a worthwhile effort. I elaborate upon this assessment below.
Incorrect Implementation of the WMO Definition
First, to hopefully put the authors somewhat at ease, I will acknowledge that this is not the first study to incorrectly implement the WMO definition. In fact, many public code packages also have erroneous implementation of the WMO definition, and it is unfortunately easy for such outcomes to occur because the language used in the 1957 WMO publication has resulted in a wide range of interpretations. When I first began to study the tropopause and apply the WMO definition ~20 years ago, I made similar mistakes which were fortunately identified by other scientists in the field and ultimately corrected before publication of my analyses. Rather than write an extensive technical description of the steps required for correct implementation, I will simply point the authors to a recent study which provides such code in multiple programming languages and discusses some important considerations for applying this definition (and the general task of identifying the tropopause for datasets of varying resolution):
Tinney et al., 2022: A Modern Approach to a Stability-Based Definition of the Tropopause, Mon. Wea. Rev., 150, 3151–3174, doi:10.1175/MWR-D-22-0174.1
I will note that the above-referenced study code includes steps for interpolation of coarse-resolution data to a finer vertical grid, which can be omitted for the high-resolution data used in this study.
Omission of Relevant Prior Research
One related and important general point I would like to make at this stage is that I noticed some glaring omissions of relevant prior work. In particular, given the focus on identification of the tropopause and its characteristics using high-resolution radiosonde data, decades of prior work with such a focus was unexplainably absent. This is especially true for the well-recognized series of early foundational studies by Birner:
Birner et al., 2002: How sharp is the tropopause at midlatitudes?, Geophys. Res. Lett., 29, doi:10.1029/2002GL015142
Birner, 2006: Fine-scale structure of the extratropical tropopause, J. Geophys. Res., 111, D04104, doi:10.1029/2005JD006301
These studies and many that follow have demonstrated well that the WMO definition, when applied appropriately, results in reliable definition of the tropopause that is insensitive to profile resolution. There are certainly failure modes of the WMO definition, but resolution is not one of them (except for resolution that is very coarse, which can of course be ameliorated by interpolation).
Another topic that I found to be poorly described and referenced relates to the statement at line 45 of the paper. In particular, the abrupt change in tropopause height near the subtropical jet streams is commonly referred to as the tropopause break. There is vast literature on the subject as well as its relation to Hadley Cell (or lack thereof, as several studies have shown that the tropopause break is not coupled to the HC). A search using the term "tropopause break" should turn up these relevant works. Additional studies that explore numerous tropical edge diagnostics to determine changes in the width of the tropics (a la Davis & Rosenlof, 2012, http://dx.doi.org/10.1175/JCL1-D-11-00127.1 and others that follow) will help resolve the HC link argument.
Specific Comments
The references for the radiosonde data sources appear to be inappropriate. In particular, the University of Wyoming page does not provide access to the full-resolution radiosonde data used in this study, but instead only includes mandatory and significant levels (as in the IGRA archive). The NOAA website referenced appears to only include surface observations and in real-time rather than a historical archive. I believe the most appropriate source of high-resolution radiosonde observations in the United States is https://www.ncei.noaa.gov/access/metadata/landing-page/bin/iso?id=gov.noaa.ncdc:C01500. There are alternative copies of the data in BUFR format from other sites, which the authors may have used. Because the data I can easily grab from U. Wyoming are on mandatory and significant levels and I do not have time to otherwise download the full-resolution BUFR data, I cannot reproduce Figure 1 from the paper to demonstrate my main point above. Nevertheless, the authors should resolve this data citation issue so that the study could be easily replicated by any reader.
Lines 74-75: I am not sure what this sentence means or implies.
Line 85: “retrieve” should be “retrieval”
Citation: https://doi.org/10.5194/egusphere-2024-4198-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 113 | 16 | 7 | 136 | 5 | 4 |
- HTML: 113
- PDF: 16
- XML: 7
- Total: 136
- BibTeX: 5
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1