the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Jul 2024
Ozone trends in homogenized Umkehr, Ozonesonde, and COH overpass records
Abstract. This study presents an updated evaluation of stratospheric ozone profile trends at Arosa/Davos/Hohenpeißenberg, Switzerland/Germany, Observatory de Haute Provence (OHP), France, Boulder, Colorado, Mauna Loa Observatory (MLO) and Hilo, Hawaii, and Lauder, New Zealand with focus on the ozone recovery period post 2000. Trends are derived using vertical ozone profiles from NOAA’s Dobson Network via the Umkehr method (with a recent new homogenization), ozonesondes, and the NOAA COHesive SBUV/OMPS satellite-based record (COH) sampled to match geographical coordinates of the ground-based stations used in this study. Analyses of long-term changes in stratospheric ozone time series were performed using the updated version (0.8.0) of the Long-term Ozone Trends and Uncertainties in the Stratosphere (LOTUS) Independent Linear Trend (ILT) regression model. This study finds a consistency of the trends derived from the different observational records, which is a key factor to the understanding of the recovery of the ozone layer after the implementation of the Montreal Protocol and its amendments that control ozone-depleting substances production and release into the atmosphere. The Northern Hemispheric Umkehr records of Aros/Davos, OHP, and MLO all show positive trends in the mid to upper stratosphere with trends peaking at ~+2 %/decade. Although the upper stratospheric ozone trends derived from COH satellite records are more positive than those detected by the Umkehr system, the agreement is within the two sigma uncertainty. Umkehr trends in the upper stratosphere at Boulder and Lauder are positive but not statistically significant, while COH trends are larger and statistically significant (within 2 sigma). In the lower stratosphere, trends derived from Umkehr and ozonesonde records are mostly negative (except for positive ozonesonde trends at OHP), however the uncertainties are quite large. Additional dynamical proxies were investigated in the LOTUS model at five ground-based sites. The use of additional proxies did not significantly change trends, but equivalent latitude reduced the uncertainty of the Umkehr and COH trends in the upper stratosphere and at higher latitudes. In lower layers, additional predictors (tropopause pressure for all stations, two extra components of Quasi-Biennial Oscillation at MLO, Arctic Oscillation at Arosa/Davos, OHP and MLO) improve the model fit and reduce trend uncertainties as seen by Umkehr and sonde.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(5260 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5260 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1821', Anonymous Referee #1, 22 Aug 2024
Review of “Ozone trends in homogenized Umkehr, Ozonesonde, and COH overpass records” by Irina Petropavlovskikh et al.
Summary and General Comments:
The authors demonstrate improvements to the LOTUS MLR ozone trends model through addition of dynamical proxies applied to ground-based (Dobson Umkehr, ozonesonde) and satellite (NOAA COH overpass and zonal) ozone datasets. Ozone trends for the lower to upper stratosphere are first presented with the “standard” Reference LOTUS model (excluding the AOD proxy) for 2000-2020, after which individual additional proxies including tropopause pressure, Eddy Heat Flux, Equivalent Latitude, etc., are added to the model to determine the best choices for a “full” model trends calculation.
The authors find that, with a few exceptions in the lower stratosphere, the trend values are mostly unchanged in the full model. However, improvements to the model adjusted R2 values and p-values indicate that the addition of proxies specifically chosen for various stations and altitudes will lead to more confidence in trend detection, as well as the possibility of detecting trends smaller in magnitude compared to a base MLR with limited proxies (i.e., as would be used with zonally averaged data).
The paper is written exceptionally well, is highly detailed, and the decision-making process for choice of additional proxies in the model and other topics are carefully explained in the text and in extensive Appendices.
I have no major concerns with this manuscript, but I do wonder if the authors explored using the Payerne ozonesonde record in addition to the Hohenpeissenberg record for the Arosa/Davos station. The Payerne record is also extremely dense, Payerne is only 50 km farther in distance from Arosa/Davos, and that is an ECC record that does not have a correction factor applied as with the Hohenpeissenberg Brewer-Mast type ozonesondes. The inclusion of Payerne sonde trends could be illuminating.
Recommendation:
I recommend publication of this paper and have only minor and technical comments below.
Specific and Line-by-Line Comments:
Line 117 and 118: It looks like there are some extra parentheses on these lines.
Line 127: SHADOZ is “Southern Hemisphere Additional Ozonesondes”
Line 133: The vertical resolution of ozonesonde data is a factor of the time response of the instrument, not altitude.
Table 1: If using the OHP ozonesonde data from NDACC, I am assuming you are using the homogenized “Corrected Ozone partial pressure” ozone values in those files, correct?
Table 2: NOAA 11 appears twice in this Table.
Line 263: I think you are missing a “m =” here
Line 362: Change “interannual” to “interannually”
Line 413: The OHP sonde and Umkehr trend differences look quite large for all Layers 3-5 (not just 4), although always within the 2 standard errors.
Figures 3 and 5: Suggest changing “Height (hPa)” to “Pressure (hPa)”
Line 528: I’m not sure I would say the results point to the inability of the model to detect non-zero trends. At this point we really don’t know and probably cannot say that trends are non-zero here in Layer 6.
Line 606 and 794: “lower” stratospheric ozone records…
Table 13: A plot similar to Figure 7 with the adjusted R2 values, but for the p-values for the Ref and Full models could be helpful and would keep the reader from having to flip back and forth between the two Tables 6 and 13.
Lines 866-869: I think it would be useful to put these questions in the intro as well (or just move them there) to very clearly motivate this study.
Line 901: Change “in case of” to “for”
Citation: https://doi.org/10.5194/egusphere-2024-1821-RC1 - AC1: 'Reply on RC1', Irina Petropavlovskikh, 22 Oct 2024
- AC2: 'Reply on RC2', Irina Petropavlovskikh, 22 Oct 2024
-
RC2: 'Comment on egusphere-2024-1821', Anonymous Referee #2, 02 Sep 2024
This paper is dedicated to improving evaluation of ozone trends using homogenized Umkehr, ozonesonde and satellite overpass data.
The use of additional dynamical proxies in the LOTUS regression model is investigated in detail, and a so-called “full model” with proxies depending on altitude and location has been applied. The authors found that the use of additional proxies did not significantly change trends but reduce trend uncertainties and improve the quality of the fit.
Please find my rather minor comments on the paper in the annotated manuscript.
- AC2: 'Reply on RC2', Irina Petropavlovskikh, 22 Oct 2024
- AC1: 'Reply on RC1', Irina Petropavlovskikh, 22 Oct 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1821', Anonymous Referee #1, 22 Aug 2024
Review of “Ozone trends in homogenized Umkehr, Ozonesonde, and COH overpass records” by Irina Petropavlovskikh et al.
Summary and General Comments:
The authors demonstrate improvements to the LOTUS MLR ozone trends model through addition of dynamical proxies applied to ground-based (Dobson Umkehr, ozonesonde) and satellite (NOAA COH overpass and zonal) ozone datasets. Ozone trends for the lower to upper stratosphere are first presented with the “standard” Reference LOTUS model (excluding the AOD proxy) for 2000-2020, after which individual additional proxies including tropopause pressure, Eddy Heat Flux, Equivalent Latitude, etc., are added to the model to determine the best choices for a “full” model trends calculation.
The authors find that, with a few exceptions in the lower stratosphere, the trend values are mostly unchanged in the full model. However, improvements to the model adjusted R2 values and p-values indicate that the addition of proxies specifically chosen for various stations and altitudes will lead to more confidence in trend detection, as well as the possibility of detecting trends smaller in magnitude compared to a base MLR with limited proxies (i.e., as would be used with zonally averaged data).
The paper is written exceptionally well, is highly detailed, and the decision-making process for choice of additional proxies in the model and other topics are carefully explained in the text and in extensive Appendices.
I have no major concerns with this manuscript, but I do wonder if the authors explored using the Payerne ozonesonde record in addition to the Hohenpeissenberg record for the Arosa/Davos station. The Payerne record is also extremely dense, Payerne is only 50 km farther in distance from Arosa/Davos, and that is an ECC record that does not have a correction factor applied as with the Hohenpeissenberg Brewer-Mast type ozonesondes. The inclusion of Payerne sonde trends could be illuminating.
Recommendation:
I recommend publication of this paper and have only minor and technical comments below.
Specific and Line-by-Line Comments:
Line 117 and 118: It looks like there are some extra parentheses on these lines.
Line 127: SHADOZ is “Southern Hemisphere Additional Ozonesondes”
Line 133: The vertical resolution of ozonesonde data is a factor of the time response of the instrument, not altitude.
Table 1: If using the OHP ozonesonde data from NDACC, I am assuming you are using the homogenized “Corrected Ozone partial pressure” ozone values in those files, correct?
Table 2: NOAA 11 appears twice in this Table.
Line 263: I think you are missing a “m =” here
Line 362: Change “interannual” to “interannually”
Line 413: The OHP sonde and Umkehr trend differences look quite large for all Layers 3-5 (not just 4), although always within the 2 standard errors.
Figures 3 and 5: Suggest changing “Height (hPa)” to “Pressure (hPa)”
Line 528: I’m not sure I would say the results point to the inability of the model to detect non-zero trends. At this point we really don’t know and probably cannot say that trends are non-zero here in Layer 6.
Line 606 and 794: “lower” stratospheric ozone records…
Table 13: A plot similar to Figure 7 with the adjusted R2 values, but for the p-values for the Ref and Full models could be helpful and would keep the reader from having to flip back and forth between the two Tables 6 and 13.
Lines 866-869: I think it would be useful to put these questions in the intro as well (or just move them there) to very clearly motivate this study.
Line 901: Change “in case of” to “for”
Citation: https://doi.org/10.5194/egusphere-2024-1821-RC1 - AC1: 'Reply on RC1', Irina Petropavlovskikh, 22 Oct 2024
- AC2: 'Reply on RC2', Irina Petropavlovskikh, 22 Oct 2024
-
RC2: 'Comment on egusphere-2024-1821', Anonymous Referee #2, 02 Sep 2024
This paper is dedicated to improving evaluation of ozone trends using homogenized Umkehr, ozonesonde and satellite overpass data.
The use of additional dynamical proxies in the LOTUS regression model is investigated in detail, and a so-called “full model” with proxies depending on altitude and location has been applied. The authors found that the use of additional proxies did not significantly change trends but reduce trend uncertainties and improve the quality of the fit.
Please find my rather minor comments on the paper in the annotated manuscript.
- AC2: 'Reply on RC2', Irina Petropavlovskikh, 22 Oct 2024
- AC1: 'Reply on RC1', Irina Petropavlovskikh, 22 Oct 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 421 | 162 | 54 | 637 | 23 | 21 |
- HTML: 421
- PDF: 162
- XML: 54
- Total: 637
- BibTeX: 23
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Irina Petropavlovskikh
Jeannette D. Wild
Kari Abromitis
Peter Effertz
Koji Miyagawa
Lawrence E. Flynn
Eliane Maillard-Barra
Robert Damadeo
Glen McConville
Bryan Johnson
Patrick Cullis
Sophie Godin-Beekmann
Gerald Ancellet
Richard Querel
Roeland Van Malderen
Daniel Zawada
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5260 KB) - Metadata XML