Study of the Lower Stratospheric Temperature over the Arabian Peninsula
Round 1
Reviewer 1 Report
OVERVIEW
The authors analyse the time series of winter mean lower stratospheric (50 hPa) temperature over the Arabian Peninsula in years 1951-2016 using the NCEP-NCAR reanalysis. No significant trends are found, although there is slight warming up to 1992 and slight cooling thereafter. They also find significant (although weak) correlations of the lower stratospheric temperature with the QBO with 1-year time lag and with the sunspot number, as well as a somewhat stronger correlation with the near-surface temperature in the Arabian Peninsula. Spectral analysis of the lower stratospheric temperature time series suggests significant periodicities at periods of 2.2, 2.54 and 13.2 years.
This is my second review of the now re-submitted paper. I appreciate the effort that the authors have made to address my previous comments and there is indeed some improvement. However, I still find the paper quite difficult to read. This is to a large part due to problems in the English language, which occasionally make it hard to follow what is actually meant. In addition, some of the methods are still incompletely described and/or poorly motivated.
Specifically, the authors should address the issues detailed in the following comments:
DETAILED COMMENTS: SUBSTANCE
1. L59-60. Dynamical processes in the troposphere lead to stratospheric cooling in the specific case that these processes make the troposphere warmer. More generally, they tend to cause opposite temperature variations in the troposphere and in the stratosphere. But the referred study [28] is on variations on the daily-to-monthly time scales and it is not clear whether the same mechanism is also significant for long-term climate change.
2. L91-93 and elsewhere. I still cannot follow the motivation of using the coefficient of variability (COV) to characterize the changes in the lower stratospheric temperature, nor the way in which the authors actually calculate COV. From the definition on L92, COV is a signal-to-noise ratio, giving the difference of the mean from its reference value (Yref) in standard deviation units. But then, on L140-141, the values of COV are -0.43 in 1961-1990, -0.42 in 1971-2000 and -0.10 in 1981-2010. How is it possible that these values are negative in all three periods? How are Yref and SRef defined? Are they always based on the full study period (1951-2016) or are different reference periods used in different cases? And what is the ultimate motivation for expressing the LST changes in terms of COV rather than directly in degrees C that would be more straightforward to interpret?
3. L172-174. Figure 1 only gives the lagged cross-correlation at one specific lag, and it is not at all “easily observed” without looking at the numeric value that the correlation is positive. What about reporting the lag correlations in a table, for example for the lags -2, -1, 0, 1 and 2, for all of the cases studied in Figs. 4, 5 and 6?
DETAILED COMMENTS: LANGUAGE ETC.
4. L35. Solomon et al. (2010) is missing from the list of references
5. L35-36. “tropospheric warming and stratospheric cooling is due to the increase in the stratospheric water vapor” should probably be “tropospheric warming and stratospheric cooling are affected by increasing stratospheric water vapor”. The increase in stratospheric water vapor is not the main cause of either stratospheric or tropospheric temperature changes. Furthermore, Manabe and Strickler (1964) certainly did not discuss the causes of recent climate change.
6. L46. polar stratospheric temperature
7. L52-53. the easterly quasi-biennial oscillation and strong solar activity tend to make the stratospheric polar vortex warmer
8. L69. (QBO) was calculated at
9. L69. Spell out PSD and give web address or reference.
10. L75-76. Eastern hemisphere: 30E and 60E.
11. L82. increased until the beginning
12. L96. “Changes in COV” or “Large values of COV”?
13. L114. whereas the time series exhibited an insignificant
14. L121. stratosphere warmed while the area close to the surface cooled
15. L127-128 and later. “significance level”, not “significant level”
16. L134. the same behaviour as in the tropics
17. L140-141. I still cannot figure out what you mean by the “average change of the LST” according to the coefficient of variability method. Do you refer to the mean values of COV in the three periods of their linear trends in these periods?
18. L177, 187 and 201. which is significant at 95% confidence level
19. L179. Between two solar minima?
20. L203-204. Of course the surface is warmer than the troposphere and the troposphere is warmer than the stratosphere (in absolute terms). Do you mean “the surface has warmed faster than the troposphere whereas the stratosphere has cooled” or something similar?
21. L211. In the extremely cold winter
22. L224 and 225. 0.05C per year
Author Response
Comments and Suggestions for Authors
OVERVIEW
The authors analyse the time series of winter mean lower stratospheric (50 hPa) temperature over the Arabian Peninsula in years 1951-2016 using the NCEP-NCAR reanalysis. No significant trends are found, although there is slight warming up to 1992 and slight cooling thereafter. They also find significant (although weak) correlations of the lower stratospheric temperature with the QBO with 1-year time lag and with the sunspot number, as well as a somewhat stronger correlation with the near-surface temperature in the Arabian Peninsula. Spectral analysis of the lower stratospheric temperature time series suggests significant periodicities at periods of 2.2, 2.54 and 13.2 years.
This is my second review of the now re-submitted paper. I appreciate the effort that the authors have made to address my previous comments and there is indeed some improvement. However, I still find the paper quite difficult to read. This is to a large part due to problems in the English language, which occasionally make it hard to follow what is actually meant. In addition, some of the methods are still incompletely described and/or poorly motivated.
Response: The English language of our paper was reviewed from the expert of the English language.
Specifically, the authors should address the issues detailed in the following comments:
Response: We are very grateful for sending the insightful comments and suggestions on our manuscript by you. The manuscript is thoroughly revised in light of your comments. The point-by-point responses for your comments and/or suggestions are as follows: -
DETAILED COMMENTS: SUBSTANCE
1. L59-60. Dynamical processes in the troposphere lead to stratospheric cooling in the specific case that these processes make the troposphere warmer. More generally, they tend to cause opposite temperature variations in the troposphere and in the stratosphere. But the referred study [28] is on variations on the daily-to-monthly time scales and it is not clear whether the same mechanism is also significant for long-term climate change.
Response: The significance of the dynamical processes in the troposphere, which lead to stratospheric cooling for long-term climate change may need to more insight from researchers.
2. L91-93 and elsewhere. I still cannot follow the motivation of using the coefficient of variability (COV) to characterize the changes in the lower stratospheric temperature, nor the way in which the authors actually calculate COV. From the definition on L92, COV is a signal-to-noise ratio, giving the difference of the mean from its reference value (Yref) in standard deviation units. But then, on L140-141, the values of COV are -0.43 in 1961-1990, -0.42 in 1971-2000 and -0.10 in 1981-2010. How is it possible that these values are negative in all three periods? How are Yref and SRef defined? Are they always based on the full study period (1951-2016) or are different reference periods used in different cases? And what is the ultimate motivation for expressing the LST changes in terms of COV rather than directly in degrees C that would be more straightforward to interpret?
Respond: The negative values are coming from “the difference between Yref and the average of total time series”.
The Yref and Sref defined based on different reference periods used in different cases.
The ultimate motivation for expressing the LST changes in terms of COV rather than directly in degrees C is “The coefficient of variation measures the reliability of the average or change in the average”. The average is less reliable when the COV is high but more reliable when the COV is low. Changes in COV show a change in the mean value.
3. L172-174. Figure 1 only gives the lagged cross-correlation at one specific lag, and it is not at all “easily observed” without looking at the numeric value that the correlation is positive. What about reporting the lag correlations in a table, for example for the lags -2, -1, 0, 1 and 2, for all of the cases studied in Figs. 4, 5 and 6?
Response: The authors think that the values of “lag correlations -2, -1, 0, 1 and 2”, for all of the cases studied in Figs. 4, 5 and 6 not needed to put in the table (the values are insignificant). However, many authors such as “Wang 2002a,b” not included for many lags in table or figure.
Wang C. 2002a. Atlantic climatic variability and its associated atmospheric circulation cells. J. Clim. 15: 1516–1536.
Wang C. 2002b. Atmospheric circulation cells associated with the El Nino–Southern Oscillation. J. Clim. 15: 399–419.
DETAILED COMMENTS: LANGUAGE ETC.
4. L35. Solomon et al. (2010) is missing from the list of references
Response: The reference “Solomon et al. (2010)” was included in the list of references
5. L35-36. “tropospheric warming and stratospheric cooling is due to the increase in the stratospheric water vapor” should probably be “tropospheric warming and stratospheric cooling are affected by increasing stratospheric water vapor”. The increase in stratospheric water vapor is not the main cause of either stratospheric or tropospheric temperature changes. Furthermore, Manabe and Strickler (1964) certainly did not discuss the causes of recent climate change.
Response: Thanks a lot for this. “tropospheric warming and stratospheric cooling are affected by increasing stratospheric water vapor”. Yes, “The increase in stratospheric water vapor is not the main cause of either stratospheric or tropospheric temperature changes”.
Of course, Manabe and Strickler (1964) certainly did not discuss the causes of recent climate change, but Solomon et al 2010 (line 35) discussed this issue.
6. L46. polar stratospheric temperature
Response: “polar stratospheric temperature” was included in the text of the revised paper
7. L52-53. the easterly quasi-biennial oscillation and strong solar activity tend to make the stratospheric polar vortex warmer
Response: “the easterly quasi-biennial oscillation and strong solar activity tend to make the stratospheric polar vortex warmer” was included instead of “the easterly quasi-biennial oscillation and strong solar activity are reasons for stratospheric polar vortex warming”
8. L69. (QBO) was calculated at
Response: “was calculated at” was included in the text of the revised paper.
9. L69. Spell out PSD and give web address or reference.
Response: Physical Sciences Division (PSD). The web address was included before.
10. L75-76. Eastern hemisphere: 30E and 60E.
Response: Thanks a lot. 30E and 60E were corrected.
11. L82. increased until the beginning
Response: “increased until the beginning” was included instead of “increased at the beginning”
12. L96. “Changes in COV” or “Large values of COV”?
Response: The authors think “Changes in COV”
13. L114. whereas the time series exhibited an insignificant
Response: “whereas the time series exhibited an insignificant” was added instead of “whereas exhibited insignificant”
14. L121. stratosphere warmed while the area close to the surface cooled
Response: “stratosphere warmed while the area close to the surface cooled” added to the text in the revised paper instead of “stratosphere has warmed while the area close to the surface has cooled”
15. L127-128 and later. “significance level”, not “significant level”
Response: “significance level” added to the text in the revised paper in L127-128 and later.
16. L134. the same behaviour as in the tropics
Response: “the same behavior as in the tropics” added into the text of the revised paper instead of “the same behavior in the tropics”
17. L140-141. I still cannot figure out what you mean by the “average change of the LST” according to the coefficient of variability method. Do you refer to the mean values of COV in the three periods of their linear trends in these periods?
Response: The authors mean by the “average change of the LST” is “the change in the average of the LST”
18. L177, 187 and 201. which is significant at 95% confidence level
Response: “which is significant at 95% confidence level” added into the text of the revised paper instead of “which significant at 95% confidence level” (L177, 187 and 201).
19. L179. Between two solar minima?
Response: Thanks. Yes, Between two solar minima.
20. L203-204. Of course the surface is warmer than the troposphere and the troposphere is warmer than the stratosphere (in absolute terms). Do you mean “the surface has warmed faster than the troposphere whereas the stratosphere has cooled” or something similar?
Response: the temperature at the surface was warmer when compared to the troposphere, and the troposphere was warmer than the stratosphere
21. L211. In the extremely cold winter
Response: “In the extremely cold winter” added into the text of the revised paper instead of “The extremely cold winter”
22. L224 and 225. 0.05C per year
Response: per year added into the text of the revised paper.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The author has satisfactorily addressed my initial concerns and the manuscript has improved. I recommend acceptance of this study for publication.
Author Response
Thanks a lot for you.
Reviewer 3 Report
Investigating the impact of the lower stratospheric temperature on regional climate of Arabian Penisolar is interesting. This study investigates the LST and ST temperature. It is found that the periodicity cycles of LST time series at periods of about 2.2, 2.54 and 13.2 years with respect to the 95% significance confidence level was found and that LST may be influenced by quasi-biennial oscillation and the sunspot cycle. The method section sounds good and the results section seems interesting. The authors have hand the comments adequately. I recommend the manuscript can be published. In the proofreading state, I do recommend the authors improve the Figure and shorten the conclusion section. For Time serials, the figures should be plotted in line-dot style.
Author Response
Thanks a lot for you.
Round 2
Reviewer 1 Report
GENERAL COMMENTS
The paper is converging to its final form. All my remaining comments are related to minor details in the presentation.
MINOR COMMENTS
L36-37. Suggested wording: … concluded that, as a global mean, the upper …
L45. delete “has”
L51. The two references ([20],[21]) should be in the end of the previous paragraph.
L107. a comparatively large standard deviation?
L111 and L113. The “2” in R2 should be a superscript
L136. 99% significance level
L177. in plural: minima or minimums
L194. ”LST changes in ST” is incomprehensible. “could contribute to the relationship between LST and ST”?
L200. coupling between the troposphere and the stratosphere
L204. delete “it is”
L215. Add “over the Arabian peninsula” to the end of the sentence.
Author Response
Response: We are very grateful for sending insightful comments on our manuscript by you. The manuscript is thoroughly revised in light of your comments. The point-by-point responses for your comments are as follows: -
Response: Our manuscript is reviewed by a native English speaking colleague.
The paper is converging to its final form. All my remaining comments are related to minor details in the presentation.
MINOR COMMENTS
L36-37. Suggested wording: … concluded that, as a global mean, the upper …
Response: “…concluded that, as a global mean, the upper…” is included in the text of the revised paper.
L45. delete “has”
Response: “has” is deleted
L51. The two references ([20],[21]) should be in the end of the previous paragraph.
Response: The two references ([20],[21]) are included at the end of the previous paragraph
L107. a comparatively large standard deviation?
Response: “a comparatively large standard deviation” is included
L111 and L113. The “2” in R2 should be a superscript
Response: The “2” in R2 is included as a superscript
L136. 99% significance level
Response: “99% significance level” is corrected
L177. in plural: minima or minimums
Response: “minima” is added instead of “minimum”
L194. ”LST changes in ST” is incomprehensible. “could contribute to the relationship between LST and ST”?
Response: ”could contribute to the relationship between LST and ST” is added
L200. coupling between the troposphere and the stratosphere
Response: “coupling between the troposphere and the stratosphere” is added
L204. delete “it is”
Response: “it is” is deleted
L215. Add “over the Arabian peninsula” to the end of the sentence.
Response: “over the Arabian peninsula” is added to the end of the sentence.
Author Response File: 
Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
OVERVIEW
The authors analyse the time series of winter mean lower stratospheric (50 hPa) temperature over the Arabian Peninsula in years 1951-2016 using the NCEP-NCAR reanalysis. No significant trends are found, although there is slight warming up to 1992 and slight cooling thereafter. They also find significant (although weak) correlations of the lower stratospheric temperature with the QBO with 1-year time lag and with the sunspot number, as well as a somewhat stronger correlation with the near-surface temperature in the Arabian Peninsula. Spectral analysis of the lower stratospheric temperature time series suggests significant periodicities at periods of 2.2, 2.54 and 13.2 years.
The study represents a straightforward time series analysis, with limited physical interpretation. It could be improved, for example, by studying how the lower stratospheric temperatures in the Arabian Peninsula are related to temperature variations at the same level in other parts of the world. Similarly, the connection between the near-surface temperature and the lower stratospheric temperature could be examined in more detail (why is the lower stratosphere cold when the surface is warm?). Nevertheless, even in its present form, the content appears sufficient for publication in Climate, although somewhat marginally so.
However, as detailed below, I nevertheless have some major concerns that should be addressed before publication.
MAJOR COMMENTS
1) The use of the NCEP/NCAR reanalysis as the only source of data might be hazardous. Several more recent and most probably more accurate reanalyses are available, particularly for the satellite era since 1979. Comparison with (e.g.) ERA-Interim or the new ERA5 for this period would strengthen the paper. Moreover, the extension of the study period back to 1951 seems problematic because very few actual observations were available at that time. In the Integrated Global Radiosonde Archive (IGRA), the first observations from the Kingdom of Saudi Arabia (at Jeddah) start in 1966 (see https://www1.ncdc.noaa.gov/pub/data/igra/igra2-station-list.txt). The paucity of observations may compromise the quality of the NCEP/NCAR reanalysis in the early part of the record, and changes in the observing system (in particular, the introduction of satellite data from ~1979) may introduce artificial trends and changes in the lower stratospheric temperature record. At the very least, these uncertainties should be honestly acknowledged in the paper.
2) More insight on the dynamics of temperature variability could be gained by comparing the lower stratospheric temperatures in the Arabian Peninsula with those in other areas. For example, a correlation map between the Arabian Peninsula LST time series and the global distribution of LST could be useful.
3) When discussing observed temperature trends, the priority must be given to the newest studies. For example, on L181-182, there are six references to studies published in 1998-2001. These are outdated! Although the IPCC AR5 assessment report from 2013 is already several years old, its Chapter 2 (https://www.ipcc.ch/site/assets/uploads/2017/09/WG1AR5_Chapter02_FINAL.pdf) is still much better up to date than any individual studies from the turn of the century.
4) The methods and data sets used should be documented more precisely. For example, the beginning of Section 2.1 states (L67-68) that “The borders of the AP are well established” but does not actually define which grid boxes were used for calculating the AP temperature time series. Similarly, there is no information on where the QBO time series comes from (or how it was calculated), and what is the source for Sunspot number time series.
5) Substantial improvement of the English language is needed. A few examples are given in the minor comments below, but this list is far from comprehensive.
MINOR COMMENTS
1. L15. What is the unit of the standard deviation? 1.3 K?
2. L31-33. It would be better to discuss the observed stratospheric temperature trends up to the 2010s based on Chapter 2 in the IPCC Working Group I Assessment report from 2013 (See Major Comment 3).
3. L32-33. I am surprised by the claim that variations in temperature are important for estimating trends in stratospheric ozone. Is it not rather the case that changes in ozone are important for explaining the trends in stratospheric temperature?
4. L40. 0.2 K / decade: positive or negative?
5. L53-55. What is the direction between cause and effect? Do the stratospheric temperatures affect the tropical Pacific sea surface temperatures or vice versa? The latter seems much more likely.
6. L76. Kingdom of Saudi Arabia.
7. L86-87. Is this not vice versa? The average should be less reliable when the standard deviation is large and, therefore, COV is small.
8. L88. It is worth noting?
9. L98. findings of the linear trend analysis?
10. L100. 1.3 °C? The units are also missing in the remainder of this section.
11. L108. ... the warmest years of the LST ...
12. L112-113. The general cooling at the surface and in the troposphere was due to reduced solar radiation (as a larger fraction of it was absorbed and reflected in the stratosphere), not due to circulation changes.
13. L114. Which years are included in the reference period?’ 1961-1990?
14. L114-116. Is the “average change” of the LST the mean anomaly of LST or its trend during the 30-year periods?
15. L116-117. “change in the average value” or just “the average value”?
16. Figure 2. Please explain the two upward pointing arrows.
17. L122. Distribution pattern à Time series. The same also applies to captions of Figs. 4-6.
18. L124. 1951-1992 (red dotted line).
19. L140-141. Anstey et al. [40] found that the QBO has a period of 35 months in a climate model simulation, but they clearly state that this is unrealistically long because the observed period is 28 months (see beginning of Section 3 in their article).
20. L145. Figure 4 illustrates the time series of ...
21. Figure 4. The lack of oscillations in the QBO time series between 1996 and 2000 seems surprising. Check your data.
22. L157. Figure 5 shows the time series of the sunspot number ...
23. L162. The time series of the sunspot number ...
24. L167. Hartley et al. 1998 = [73]?
25. L180-182. These references are too old for discussion of observed temperature trends. See Major comment 3.
26. L199-200. Why should one use COV for deciding how LST changes, rather than directly comparing the values of LST between different periods?
27. L200. This should probably be just “the average value of LST tends to increase”. Saying that a change tends to increase is repetitive, because the word “increase” already incorporates the change.
28. L209. -0.50 at zero lag.
29. L210-212. The negative correlation on the interannual time scale (if there is any, after removing the trends?) cannot be explained by greenhouse gases, because there is no significant interannual variability in greenhouse gas concentrations. Thus, it must be caused by some other mechanism.
Reviewer 2 Report
Review of “Study of the lower Stratospheric Temperature over the Arabian Peninsula” by Hasanean and Labban.
This manuscript analyzes lower stratospheric temperature averaged over the Arabian Peninsula region. The authors examined the time series, conducted spectral analysis and compared with a few indices to explore the physical mechanism. This manuscript has some merits, although the analysis is quite simple. However, the English writing is poor, and I strongly suggest the authors to have professional English speakers help improve it. I have listed my concerns/suggestions below.
1. L12: “related” => “relation”.
2. L59-60: There’s no section 6, and section 4 is summary section.
3. L77: the authors separate their analysis into two period, yet the reason why they pick the year 1992 is not clear. Please explain this explicitly in the manuscript.
4. L85-87: this parameter is related to both mean values and standard deviation. A small value in the parameter could be due to small numerator and/or large denominator. I wonder if this affects the conclusion about changes in the mean value, which would neglect changes in the standard deviation (e.g. L116-117).
5. L98: “findings” => “findings using”
6. L131: “substantial” => “significant”
7. Figs 4-6: this part of simple correlation analysis might provide some insights into the physical connections, but it is important to test whether the correlation coefficients that the authors have obtained are statistically significant or not.
Reviewer 3 Report
This paper examines the lower stratospheric temperature at 50hPa (LST) in winter and its related with surface air temperature (ST) over Arabian Peninsula (AP) in the period (1951 – 2016). The manuscript needs substantial revision before considering publication. The writing in most cases is not clear.
(1) the manuscript contains too many abbreviations, which may the readability of the manuscript bad.
(2) the orignization of the paper is not good. Statement in Line 57-60 should be deleted.
(3) The unit and title for the y-axis in Figure 2 is missed. The figure is thus unreadable.
(4) Figure 3 is unreadable as well. It is recommended the authors replot the figure.
(5) Figure 4-5 is similar as Figure2. The Y-axis is missed for these double Y-axis plot. Only the authors know what the figure means.
At this state, the reviewer believe the manuscript is unsuitable to publish in CLIMATE.
