Trends in Extreme Precipitation Indices in Northwest Ethiopia: Comparative Analysis Using the Mann–Kendall and Innovative Trend Analysis Methods

Round 1

Reviewer 1 Report

This study looks at trends in precipitation extremes in three small regions (~50km scale) of Ethiopia during the period from 1981 to 2018. Gridded data, which I understand to be daily precipitation, is provided by the national weather service. Extremes are calculated using the RClimDex R package while MK and ITA trend analysis is done with trendchange R package.  

 

The paper is well structured and very clear from the start. The abstract spells out exactly what the authors want to accomplish in the paper. Results are presented neatly and concisely in the summarizing tables and the state-of-the-art is exhaustively summarized and appropriately referenced in the bibliography.

While I see no formal errors or significant holes in the study, possibly some further description of the precipitation dataset would be helpful, I think this paper may not make quite enough of an impact to warrant publication. I am not surprised that many, or even most, of the extreme precipitation indexes present significant trends over the last 40 years. Some are positive trends, and some are negative trends. Not all these trends are consistent with the Clausius-Clapeyron equation and a warming atmosphere (the authors refer to this issue in lines 29 to 31).

 

The authors do not offer a possible physical explanation for their emerging results. Nor do they try to explain why some indexes show significant trends in one or two regions but not all three. Could it be the type of terrain, exposure of the slopes, changes in circulation, differences in elevation? To be clear, I did not expect the trends to all be significant or even have the same sign. In fact, I suspect that these regions are so small that they are at the limit of what can produce a significant climate signal over 40 years. Of course, if one tests 10 indexes in 3 regions with 2 methods, one has 60 separate chances of getting a significant signal. This is why I think that the presentation of results alone, without a suggested underlying explanation, is not as meaningful as one would want.

 

Similarly, I do not think it is a novel result that ITA trend detection is more effective than MK when applied to non-monotonic trends. In fact, I believe that ITA was developed in part because of the limitation of MK in the presence of non-monotonic trends. The authors are clearly experts on this matter since they refer to it in detail and refer to the relevant articles as well.   

 

For theses reasons I suggest to the editor that the authors be asked to extend their study to include further analysis and reflections, including possible physical explanations, to account for both negative and positive trends. Alternatively, they could add more specifics as to the use of their results. In line 363, the authors state: ” the results provided useful information on extreme precipitation events in the study area”. Could the authors expand on the matter? Who will be using these results? Policy makers? Water resource managers? 

 

On a different topic, there is no description of the data set used. It would be interesting to know:

·         How many stations were used and their location.

·         What were the data gaps in the station data?

·         Is the data available to others?

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Dear Authors,

I find the manuscript interesting, there are an overall coherence and relation to the scope of publication in the journal, but the paper is average-written and low-argued and several aspects must be improved. I think your paper at present is not ready for publication.

Please make sure that there is a scientific objective in your manuscript. This is not a novelty. What is the main aim novelty present in this study is not clear. The manuscript must include: What is new in your investigation?

Abstract

Abstract section need to be mentioned the source of data used in this study.

L 13-16 “Based on the ITA result, the calculated values of nine indices (90% of the analyzed indices) showed significant increasing trends (p<0.01) in Lay Gayint. In Tach Gayint, 80% (eight 14 indices) showed significantly increasing trends at p >< 0.01) in Lay Gayint. In Tach Gayint, 80% (eight (seven?) indices) showed significantly increasing trends at p <0.01. On the other hand, 50% (60%?) of the extreme indices showed significant downward trends (p<0.01) in Simada. ><0.01) in Simada.”

Please the results given above should be checked.

L 17-18 “In Tach Gayint, 40% of the extreme indices showed significant increasing trends at p<0.01 (0.05?).

Please the result given above should be checked.

Introduction

It is bad way to start the introduction. Revised the first paragraph with some better part, which impact the research and attract the readers.

L 81-82 “…previous studies have shown that the presence of autocorrelation could affect trend identification using the MK and Sen’s slope methods…” As is known, the MK and MMK methods are based on many assumptions. Therefore, the use of these methods and the results obtained are considered valid according to these assumptions. On the other hand, there are no acceptance and statistical significance levels in the Sen’s Slope method.   According to the literature review, the classical Mann Kendall (MK) (Mann, 1945; Kendal, 1975) and Sen’s slope (SS) (Sen, 1968) methods are frequently used to determine trend directions and magnitudes. MK and SS methodologies use the whole time series and do not make any categorical distinction, say, among “low”, “medium” and “high” values in search of partial trends. They search for temporal monotonic trend component mostly in holistic and monotonic manners without any distinction between “low”, “medium” and “high” values, which may have different trend patterns. Classical trend tests can be criticized because of the above-mentioned deficiency. Additionally, all the classical trend determination methodologies try to find holistic monotonic trend either over the whole record period or on pieces of sub-periods (Şen, 2017). Using the acceptance and non-acceptance-based methods together and evaluating their results with experience has undoubtedly more advantages to obtain correct trend inferences. In the case of using methodologies such as MK and Spearman Rho (SR), methods with a large number of assumptions, tests should be made on the assumptions before the use of the method, and the results should be interpreted and used within the framework of the method assumptions. In the case of using two different methods, the coherency relationship between the results should be evaluated by interpreting them logically. In summary, the results of the classical MK and SS methodologies do not support each other in some cases. The introduction and methodology should be updated with the help of new references with involve ITA methods. The following article can be reviewed in order to be supported for the manuscript.

·       Birpınar, M.E., Kızılöz, B. & Şişman, E. Classic trend analysis methods’ paradoxical results and innovative trend analysis methodology with percentile ranges. Theor Appl Climatol (2023). https://doi.org/10.1007/s00704-023-04449-6

Materials and methods

Methods are clearly presented

All writing on the figure 1 must be legible. The color of the study districts’s legend on Figure 1 should match the color used on the map.

English language and style are fine/minor spell check required. For example; Line 153 “……..the Sens Slope” is revised as the Sen’s Slope.

L154 “…the magnitude (true slope) of the trend….” What does true slope mean? It is not suitable to use in this way.

The Authors may avoid to multiply references in paranthesis. Each reference can be justified by it is used and at least short assessment provided. E.g Line 90 “…method [2,40–42].”, Line 156 “…can be found in [22,58,59].”

The Graphical Innovative Trend Assessment (G-ITA) method is employed (Şen 2012) to assess temperature and precipitation changes. Therefore, Line 164 Reference should revised as [39].

L 168-169  “…….therefore, the first 19 observations were placed in the first half and the next 19 observations in the second half of the time series” This statement is wrong. It should be revised in accordance with the methodology.

Before abbreviations are used, they should be written clearly in the manuscript where they are first used. E.g in table 1, Line 131 RH test etc.

Flow chart should be added for methodologies.

Result

Line 206 “The trend in precipitation extremes from 1981 to 2018 as recorded by the MK/MMK test, is summarized in Table 2”  MK and MMK test results are expected to be different from each other. If it is the same, it is sufficient to use MK. If different results are obtained, the MMK test should be preferred.

Table 2 has irregularity in shape. It should be checked.

There are many statements in this section that are inconsistent with the tables. It should be corrected. e.g Line 208 “The rate of change in Tach Gayint was 0.03 mm/day (p<0.01)”, Line 216-217 “Results showed a statistically significant decreasing SDII trend in Lay Gayint (0.05 mm/day)”

Discussion and Conclusion

How would your study help future investigations? All this should be presented in more detail.

Conclusion section needs to be more scientifically written.

As a result, I request authors to share the data. I wish to check some part.

Sincerely,

 

 

 

 Minor editing of English language required

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have addressed all of the concernes I raised in my earlier review.

I have no remaining objections to the publication of this well written paper. 

 

Reviewer 3 Report

Dear Authors,

The only minor point is to correct the text in lines 105,108, 114, 233, 245, 256, Sen should be changed to Şen. The words where are wrong.

Sincerely,