Comparison of meteor radar and satellite winds in the Brazilian equatorial region

Paulino, Ana Roberta; Rodrigues, Delis Otildes; Paulino, Igo; Lima, Lourivaldo Mota; Buriti, Ricardo Arlen; Batista, Paulo Prado; Ridley, Aaron; Wu, Chen

doi:https://doi.org/10.5194/angeo-43-183-2025

Articles | Volume 43, issue 1

https://doi.org/10.5194/angeo-43-183-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/angeo-43-183-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 43, issue 1

Regular paper

|

06 Mar 2025

Regular paper |

| 06 Mar 2025

Comparison of meteor radar and satellite winds in the Brazilian equatorial region

Ana Roberta Paulino, Delis Otildes Rodrigues, Igo Paulino, Lourivaldo Mota Lima, Ricardo Arlen Buriti, Paulo Prado Batista, Aaron Ridley, and Chen Wu

Download

Final revised paper (published on 06 Mar 2025)
Preprint (discussion started on 18 Jul 2023)

Interactive discussion

Status: closed

RC1:
'Comment on angeo-2023-23', Anonymous Referee #1, 18 Aug 2023
This paper aims at investigating the horizontal (zonal and meridional) wind velocity at about 80-100 km altitude observed with the meteor radar at Sao Joao dos Cariri, Brazil and TIDI on board the TIMED satellite. I appreciate the devoted efforts by the authors to investigate characteristics of both ground-based radar and satellite measurements of the wind velocity in the mesosphere and lower thermosphere (MLT) region, where not so many effective wind velocity observations are available. However, I have major concerns regarding the statistical treatments on the comparison of the wind velocity data between the two techniques, as well as description and interpretation of the analyzed results.
The authors carried out three types of comparisons, which are (a) snap-shot profiles on a specific day (15 March 2006), (b) monthly mean of hourly wind velocity, and (c) day-to-day variations of wind velocity at 12 UT for one full year in 2006.
Regarding (a), discrepancies between the meteor radar and TIDI are discussed, attributing to the difference in the field of view, the integration time of each measurement and the effects of gravity waves. However, I particularly do not agree that gravity waves significantly change the wave structure within four minutes, which are as short as the buoyancy period, but spatial separation between the two TIDI profiles may explain the discrepancy. Please show the horizontal extent of the meteor radar illuminating area and the locations of the TIDI measurements. In any case, single snap-shot is not enough to draw a conclusive remark. As the authors stated on Line 78, comparisons should be repeated for more cases.

For the comparison (b), it is explained as the “daily mean wind”, but the figures seem to show the (local) time variations of the hourly wind velocity averaged over one month. Figure 2 with the meteor radar clearly shows the behavior of the diurnal and semidiurnal tides. On the other hand, Fig. 3 with TIDI includes irregular variations. Are the number of data comparable between the two determinations? I suspect Fig. 3 does not sufficiently smear out the waves other than tides, such as gravity waves, equatorial waves and so on, which are not synchronous to the local time. The same comments apply to Figs. 4 and 5.

The day-to-day variations of the wind velocity at 12 UT are shown in Figs. 6 and 7 at 90 km in year 2006. Because the one hour data at 12 UT is crucially affected by tides as recognized in Figs. 2 and 4, it does not represent the daily mean wind velocity. Instead, the wind velocity averaged throughout one full day should be used. Although the variations are explained in term of AO (annual oscillation), SAO (semiannual oscillation) and intra-seasonal oscillations, I do not clearly recognize signals of AO, SAO and others. At least, a harmonic analysis should be applied to identify AO and SAO. Statistical comparison over one full year, assuming the Gaussian distribution, may not be enough to show consistency between the two techniques. Day-to-day variations of the wind velocity and the decomposed AO and SAO signals should be statistically tested.

Overall, description and interpretation of the results are not fully convincing. Although the authors referred to earlier studies on related subjects, little discussions are given on agreement, discrepancy and progress compared to these published results.
I hope this study will be considerably improved, provided the authors revise the data analysis procedures and investigation of the results. However, I am sorry for not becoming positive to recommend publication of this manuscript in Annales Geophysicae.
I am afraid I do not clearly understand some statements. I would like to recommend the manuscript be fully refined considering the specific comments listed below:
L 9: “vertical wind profiles” is misleading.
L 16: Are “the acoustic waves” recognized evident in the MLT region?
L 25-26: “satellite measurement of wind” and “using wind measurements” are redundant. In addition, “wind” should be reworded as “wind velocity” everywhere.
L 44: Is “transceiver” commonly used to explain configuration of a radar?
L 48: Why is the meteor echo rate so variable between 1,000 and 3,000? I am also interested in the local time dependence of the meteor echo rate.
L 49-: Description of a meteor radar system could be refined.
L 53: Does “mean wind” indicate the determination of the horizontal wind velocity in 4 km x 1 hour? How many meteor echoes in each bin?
L 62: Does “vertical measurement” mean “measurement of a vertical (height) profile”? Please refine the sentence.
Figure 1 and L 76-77, L79-80: Comparison of snap-shot profiles is not enough to derive the conclusive statements. If comparisons for more cases are expected to show a better agreement, please extend the analysis, and show the results.
L 80-84: I am curious why the ducting of gravity waves is explained in detail. Are their effects important in discussing the comparisons?
L 88: Why is TIDI advantageous to estimate Ri?
L 89-90: I simply do not understand this statement.
L 92: “climatological” is not an appropriate wording, as just one year data is analyzed.
Figure 2 and L 94: Fig. 2 does not seem to show “daily mean winds”, but the local time (diurnal) variations of hourly wind velocity averaged over one month.
Figure 2 and L 99: Because large amount of the meteor radar data is averaged, short period waves are smeared out, which do not synchronize with local time. Such averaging is not completely achieved for TIDI, so that irregular patterns appear in Fig. 3. Thus, the discrepancy is simply attributed to the amount of available data for TIDI, right? Please show the number of TIDI data used for comparisons.
L 105-106: How does “mask the vertical propagation of tides” mean? Are the small ooscillations (gravity waves) interacting with tides? Or, do they just visually overwrite the regular progression pattern of tides? Why can the “short period structures” be captured after averaging over 60 days, which do not seem to synchronize with local time?
L 108-109: I do not understand this statement. Is the accuracy of TIDI dependent of latitudes?
Figures 4 and 5: The same concerns as in Figs. 2 and 3.
L 119-120: What is the “noise” of TIDI?
Figures 6 and 7: If the long-term variations, such as AO and SAO, are the target, daily mean wind is appropriate, rather than one hour data at 12 UT.
Figure 6 and L 130: I do not clearly see AO with maximum in summer. How are the intra-seasonal oscillations detected?
L 131: “even the zonal wind” Is the zonal wind shown in Fig. 6? Am I missing anything?
Table 1: The Gaussian distribution over one full year is not very useful. Day-to-day variations of the wind velocity and the AO and SAO signals should be statistically tested.
Section 4: Reviewing descriptions and explanations in Section 3, I am afraid I am not convinced with the concluding statements in this section.
More appropriate references could be cited on earlier studies about the atmosphere dynamics and the measurement techniques. Some of self-citations to the author’s group do not seem inevitable.
Citation: https://doi.org/10.5194/angeo-2023-23-RC1
- AC1: 'Reply on RC1', Ana Roberta Paulino, 20 Dec 2023
  
  Referee: This paper aims at investigating the horizontal (zonal and
  
  meridional) wind velocity at about 80-100 km altitude observed with
  
  the meteor radar at Sao Joao dos Cariri, Brazil and TIDI on board the
  
  TIMED satellite. I appreciate the devoted efforts by the authors to
  
  investigate characteristics of both ground-based radar and satellite
  
  measurements of the wind velocity in the mesosphere and lower
  
  thermosphere (MLT) region, where not so many effective wind velocity
  
  observations are available. However, I have major concerns regarding
  
  the statistical treatments on the comparison of the wind velocity data
  
  between the two techniques, as well as description and interpretation
  
  of the analyzed results
  
  Authors: We appreciate the time which was dedicated to revise our
  
  manuscripts and the important contributions pointed out by the Referee
  
  #1.
  
  Referee: Regarding (a), discrepancies between the meteor radar and
  
  TIDI are discussed, attributing to the difference in the field of
  
  view, the integration time of each measurement and the effects of
  
  gravity waves. However, I particularly do not agree that gravity waves
  
  significantly change the wave structure within four minutes, which are
  
  as short as the buoyancy period, but spatial separation between the
  
  two TIDI profiles may explain the discrepancy. Please show the
  
  horizontal extent of the meteor radar illuminating area and the
  
  locations of the TIDI measurements. In any case, single snap-shot is
  
  not enough to draw a conclusive remark. As the authors stated on Line
  
  78, comparisons should be repeated for more cases.
  
  Authors: The referee is right! We are revising this part of the
  
  manuscript. We are analysing other periods comparing the
  
  measurements, which will allow us to improve our interpretations. In
  
  the revised version of the manuscript we will revise carefully this
  
  point considering the contribution of the Referee.
  
  Referee: For the comparison (b), it is explained as the “daily mean
  
  wind”, but the figures seem to show the (local) time variations of the
  
  hourly wind velocity averaged over one month. Figure 2 with the meteor
  
  radar clearly shows the behavior of the diurnal and semidiurnal tides.
  
  On the other hand, Fig. 3 with TIDI includes irregular variations. Are
  
  the number of data comparable between the two determinations? I
  
  suspect Fig. 3 does not sufficiently smear out the waves other than
  
  tides, such as gravity waves, equatorial waves and so on, which are
  
  not synchronous to the local time. The same comments apply to Figs. 4
  
  and 5.
  
  Authors: The referee is correct! It is a composite day plot of the
  
  hourly mean winds. We have revised this part of the manuscript. The
  
  main discrepancies between Figures 2 and 3 is that: for the satellite
  
  measurement we have, on average four passages of the satellite during
  
  a day, which is not sun synchronous and along the two months can cover
  
  a day, while the meteor radar presents data every hour. We can not
  
  directly determine gravity waves from the measurements in Figures 3
  
  and 5. What we meant was that, the meteor radar averages out the
  
  effect of the gravity waves while the satellite measurements present
  
  more variability. We have revised the statements on these statements.
  
  Thank you for these contributions.
  
  Referee: The day-to-day variations of the wind velocity at 12 UT are
  
  shown in Figs. 6 and 7 at 90 km in year 2006. Because the one hour
  
  data at 12 UT is crucially affected by tides as recognized in Figs. 2
  
  and 4, it does not represent the daily mean wind velocity. Instead,
  
  the wind velocity averaged throughout one full day should be used.
  
  Although the variations are explained in term of AO (annual
  
  oscillation), SAO (semiannual oscillation) and intra-seasonal
  
  oscillations, I do not clearly recognize signals of AO, SAO and
  
  others. At least, a harmonic analysis should be applied to identify AO
  
  and SAO. Statistical comparison over one full year, assuming the
  
  Gaussian distribution, may not be enough to show consistency between
  
  the two techniques. Day-to-day variations of the wind velocity and the
  
  decomposed AO and SAO signals should be statistically tested.
  
  Authors: It is a good suggestion from the Referee and we can implement
  
  it for the revised version. However, it is important to have in mind
  
  that there are only a few soundings of the satellites during the day.
  
  In the revised version, we will consider this suggestion and improve
  
  the Discussion based on it. Additionally, we can provide a spectral
  
  analysis to show the AO and SAO and test them statistically, as
  
  suggested.
  
  Referee: Overall, description and interpretation of the results are
  
  not fully convincing. Although the authors referred to earlier studies
  
  on related subjects, little discussions are given on agreement,
  
  discrepancy and progress compared to these published results.
  
  Authors: We hope that the previous suggestions of the Referee #1 can
  
  help us to improve our discussion. Unfortunately, there are few
  
  publications on this topic to support our interpretations.
  
  Referee: I hope this study will be considerably improved, provided the
  
  authors revise the data analysis procedures and investigation of the
  
  results. However, I am sorry for not becoming positive to recommend
  
  publication of this manuscript in Annales Geophysicae.
  
  Authors: We agree with the main concerns from Referee #1 and
  
  certainly, after the revision, the referee will be more comfortable to
  
  recommend our paper for publication.
  
  Referee: I am afraid I do not clearly understand some statements. I
  
  would like to recommend the manuscript be fully refined considering
  
  the specific comments listed below:
  
  Authors: Thank you for the contribution reading carefully the
  
  statements and suggestion improvements. We have revised all statements
  
  according to the recommendations.
  
  Citation: https://doi.org/10.5194/angeo-2023-23-AC1
RC2:
'Comment on angeo-2023-23', Anonymous Referee #2, 06 Nov 2023

This manuscript presents a comparison between mesospheric winds obtained by a meteor radar located in north east Brazil (Sao Joao do Cariri), and the TIDI instrument onboard the TIMED satellite. The authors looked at almost coincident measurements, but also longer term observations. While the coincident measurements show a large variability and significant differences (even between 2 consecutive TIDI values), monthly averages and large-scale oscillations such as SAO, are in some kind of agreement.

This paper is correctly written and but there is nothing really new at this stage. Similar comparisons have been done between radar wind measurements and satellite observations, as mentioned in the text (Xu et al., 2009, John et al., 2011, Su et al., 2014).
Main comments:

- Is it possible to have a map showing the coverage of both instruments (and at two times for TIDI), especially for Figure 1? There is a large discrepancy between two consecutive TIDI measurements, so it would be interesting to see if these measurements were made over mostly different regions.

- According to what you wrote (l. 72-74), because of the ways the two instruments work and what regions of the atmosphere they measure, it seems unlikely coincident measurements will be similar, so I'm not sure this part of the investigation is meaningful.

- l. 107: how many satellite soundings were used to plot the monthly mean winds? Did you use a threshold to limit the small sampling effect? Maybe you should.

- For Figure 6, you could have used the meteor data closest to the TIDI measurements instead of at 12 UT every day, especially considering the short-term variability in TIDI values, and the effects of tides on the winds.

- l. 135: why should you consider that the measurements follow a Gaussian distribution? Instead, could you fit AO and SAO oscillations to both curves. It would be easier to compare them.

- l. 150: not sure TIDI is reliable to study short scale GW if there is so much variability in 3 min (Figure 1).

Minor edits:

l. 5-6: ...which have a vertical resolution of 2.5 km, starting...

l. 9-10: While TIDI...

l. 13: for both techniques

l. 18: remove "of the atmosphere"

l. 26: others

l. 29: oscillations

l. 32: hand, satellite measurements

l. 40: a meteor radar

l. 47: to the ground

l. 54: wind components

l. 59: TIDI...using airglow emissions as...

l. 60: airglow emissions. It has a...

l. 62: there are vertical...

l. 73: ... hand, TIDI...

l. 83-84: remove "in the MLT"

l. 85: their interaction with...

l. 87: ratio

l. 88: TIDI

l. 89-90: can you clarify this sentence?

l. 98: upper levels

l. 101-102: TIDI

l. 102: within 60 days

l. 105: presented well-defined oscillations

l. 106-107: oscillations during the day could modulate the observed diurnal tide phase.

l. 111: smaller

l. 113: as Figure 3.

l. 114: zonal winds are larger than the meteor radar ones...period

l. 115: favorably

l. 119: remove "in the wind"

l. 120: origined

l. 127: during 2006

l. 131: Even though... presented

l. 134: oscillations

l. 135: Figures 6 and 7 obey...

l. 137: close to each other

l. 138: the values of the two measurements... close, could obey...

l. 148: TIDI

l. 155: smoother

l. 159: studies long term dynamics in the MLT.
Figure 1: switch the times (14:13 and 14:17) so they are chronological. Caption: ... winds measured...

Figure 4: using

Citation: https://doi.org/10.5194/angeo-2023-23-RC2
- AC2: 'Reply on RC2', Ana Roberta Paulino, 20 Dec 2023
  
  Referee: This manuscript presents a comparison between mesospheric
  
  winds obtained by a meteor radar located in north east Brazil (Sao
  
  Joao do Cariri), and the TIDI instrument onboard the TIMED satellite.
  
  The authors looked at almost coincident measurements, but also longer
  
  term observations. While the coincident measurements show a large
  
  variability and significant differences (even between 2 consecutive
  
  TIDI values), monthly averages and large-scale oscillations such as
  
  SAO, are in some kind of agreement. This paper is correctly written
  
  and but there is nothing really new at this stage. Similar comparisons
  
  have been done between radar wind measurements and satellite
  
  observations, as mentioned in the text (Xu et al., 2009, John et al.,
  
  2011, Su et al., 2014).
  
  Authors: Thank you for your time revising our manuscript and for the
  
  important contributions suggested for our manuscripts. Even
  
  considering the previous studies on this topic comparing these kinds
  
  of observations, we have pointed out unique aspects and our
  
  conclusions are important to understanding the variability of the
  
  horizontal wind in the equatorial region. We are going to implement
  
  all suggestions from Referee #2 in order to improve the quality of the
  
  presentation of the manuscript and the interpretations.
  
  Referee: Is it possible to have a map showing the coverage of both
  
  instruments (and at two times for TIDI), especially for Figure 1?
  
  There is a large discrepancy between two consecutive TIDI
  
  measurements, so it would be interesting to see if these measurements
  
  were made over mostly different regions.
  
  Authors: Yes, it is. We are going to add an extra Figure in the
  
  revised version showing the coverage of both instruments. Regarding
  
  the comparisons of Figure 1, we are going to improve it including
  
  further analysis and improving the interpretation as suggested by the
  
  Referee #1 as well.
  
  Referee: According to what you wrote (l. 72-74), because of the ways
  
  the two instruments work and what regions of the atmosphere they
  
  measure, it seems unlikely coincident measurements will be similar, so
  
  I'm not sure this part of the investigation is meaningful.
  
  Authors: Yes, we agree, however, the importance of this comparison is
  
  to show that the horizontal wind has a large short time variability.
  
  Referee: l. 107: how many satellite soundings were used to plot the
  
  monthly mean winds? Did you use a threshold to limit the small
  
  sampling effect? Maybe you should.
  
  Authors: At least two measurements within an hour interval. There was
  
  no threshold, blank spaces represent intervals without measurements.
  
  We need to have in mind that the satellite crosses the São João do
  
  Cariri area only one or two times in a day and the measurements are
  
  not sun synchronous.
  
  Referee: For Figure 6, you could have used the meteor data closest to
  
  the TIDI measurements instead of at 12 UT every day, especially
  
  considering the short-term variability in TIDI values, and the effects
  
  of tides on the winds.
  
  Authors: Thank you for the suggestion! Yes, we can do that or use a
  
  daily average as suggested by the Referee #1. We are performing some
  
  tests and in the final version, we are going to present this Figure
  
  improved.
  
  Referee: l. 135: why should you consider that the measurements follow
  
  a Gaussian distribution? Instead, could you fit AO and SAO
  
  oscillations to both curves. It would be easier to compare them.
  
  Authors: Thank you for the suggestion, which is in agreement with the
  
  suggestion of Referee #1 as well. We are going to do that in the
  
  revised version of the manuscript.
  
  Referee: l. 150: not sure TIDI is reliable to study short scale GW if
  
  there is so much variability in 3 min (Figure 1).
  
  Authors: We think the statement was not clear enough. We meant that it
  
  is important to support studies involving gravity waves, for example
  
  to evaluate refraction, reflection and absorption of gravity waves. We
  
  are going to revise this statement.
  
  Referee: Minor edits...
  
  Authors: We appreciate all minor suggestions and in the revised
  
  version, all of them will be fixed.
  
  Citation: https://doi.org/10.5194/angeo-2023-23-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (30 Dec 2023) by Gunter Stober

AR by Ana Roberta Paulino on behalf of the Authors (08 Apr 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (09 Apr 2024) by Gunter Stober

RR by Anonymous Referee #3 (09 May 2024)

RR by Anonymous Referee #4 (08 Jul 2024)

ED: Reconsider after major revisions (further review by editor and referees) (08 Jul 2024) by Gunter Stober

AR by Ana Roberta Paulino on behalf of the Authors (22 Jul 2024) Author's response Author's tracked changes Manuscript

ED: Publish subject to revisions (further review by editor and referees) (22 Aug 2024) by Gunter Stober

AR by Ana Roberta Paulino on behalf of the Authors (22 Aug 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (26 Aug 2024) by Gunter Stober

RR by Anonymous Referee #3 (02 Sep 2024)

RR by Anonymous Referee #4 (27 Sep 2024)

ED: Publish subject to minor revisions (review by editor) (27 Sep 2024) by Gunter Stober

AR by Ana Roberta Paulino on behalf of the Authors (25 Oct 2024) Author's response Author's tracked changes Manuscript

ED: Publish as is (20 Dec 2024) by Gunter Stober

AR by Ana Roberta Paulino on behalf of the Authors (01 Jan 2025)

Short summary

Comparisons of wind measurements using two different techniques (ground-based radar and satellite) in Brazil during 2006 were made in order to point out the advantages of each instrument for studies in the mesosphere and upper thermosphere. (i) For short-period variations, the measurements of the satellite were more advantageous. (ii) The monthly climatology using the radar was more appropriate. (iii) For long periods (longer than a few months), both instruments responded satisfactorily.

Comparison of meteor radar and satellite winds in the Brazilian equatorial region

Download

Interactive discussion

Peer review completion

Suggestions for revision or reasons for rejection

Suggestions for revision or reasons for rejection

Suggestions for revision or reasons for rejection