Analysis of Quadratic Correlation between Dryness Indices and Wine Grape Yield to Estimate Future Climate Impacts in Hungary

Round 1

Reviewer 1 Report

    The subject is relevant and interesting. Nevertheless, some critical issues introduced massive uncertainty to the current results as:

1- Climate change research is generally a long time series and too few samples will reduce the credibility of the results. A concept exists in regression analysis modeling that the number of samples is more than 10 times for the independent variable. The number of samples in Figure 14 and Figure 15 is only 5, which will reduce the reliability of the results.

2- The author's description for the study area is not clear enough. Are the 22 wine grape regions concentrated in the red block of the Figure 1? The research results cannot represent the overall situation of Hungary when the study region is too small.

Minor issues:

1. L36 : Is wine regions or wine grape region？

2. L72: Is the third level title for 2.2.3 clerical error?

3. L89：The right bracket symbol is missing.

4. Irrigation is an important factor for the results, but irrigation data was not mentioned in the paper. It is understood that the laws of some European countries prohibit to irrigate for vineyards. What is the irrigation situation of vineyards in the study area need to explain.

5. L152：Similar to comment 1. Is the wine region and wine grape region in the same place?

6. L166-L68：The author mentioned here that only the warm part of the year with a daily mean temperature above 10°C are involved to calculate the dryness indices. What is the basis?

7. L181：The source of extraterrestrial radiation data shall be given.

8. L232：Similar to comment 1. Is wine yield or wine grape yield？

9. L252-L254：It is mentioned in the paper that the influence for non-meteorological factors can be filtered by processing for the linear trend deviation. Corresponding references shall be provided. Nicholl suggest that to remove the effects of the trends in the yield and climate time-series by calculating first differences of the variables (that is, year-to-year changes) and then calculating the relationships between these first differences. This approach avoids the confounding influence of long-term variations such as changes in crop management.

-Nicholls, N., 1997. Increased Australian wheat yield due to recent climate trends. Nature 387 (6632), 484-485.

10.  The method for analysis of quadratic correlation is the core method of the paper and the author should provide the explanation for the method.

Comments for author File: Comments.docx

Author Response

Dear Reviewer 1, please find attached our response in a separate file.

Author Response File: Author Response.docx

Reviewer 2 Report

This paper need very major revision.

The most crucial question- is- why with so small number of points authors decided to use only one type of the fits... Based on data- sometimes one can see linearity... 

The second problem is that I cant find a section with validation of the results of the model... without that it is hard to say that this model is correct.

It is also not clear wow the authors estimated NDD, CDDm CWB, DI, VIW for future? There is also no plots which could presented the data in the past...

The process of the creation of the model and results should be better presented.

 

Other comments: Please remove the abbreviations from abstract

Line 73-132: Please add the equations to the indices that you used with explanation of the parameters that would be used in formula

Line 135:add reference to FORESEE database

Line 136: add which variables

Line 141: add some more information about the RCM not only GCM and Institution (some references can be not accessible for audience) so just few words about each models would be crucial

ETo- add the potential source of this data

Line 191: the parameters should be with indices in formula when you use monthly data

Line 223: add reference to the indicator

Can you present the variation of the factors that have the impact on VWI on any plot, how day changes year-to-year or month-to- month? Maximum and minimum values.

If data for 2005-2021 are unknown, and known are data for the period since 2017-2021 why the simulation is not linked to these periods? Was a period 2017-2021 validated by model?

In general it is not clear- how the periods 1986-2005, 2016-2036, 2081-2100 were choosen and what was a base for this devitions?

Figure 13 and 14, 15: why do you used this type of trend lines? Not other type? What is a purpose to not to use other type of regression ? Some plots seems to present linear dependences.

In general the numbers next to axis are unreadable. This figure is unreadable.

Table 9 Add information what Rate of significance is. It should be clear for audience without reading details in the text

Table 10,11- add explanation of each parameters presented in the table. It should be clear for audience without reading details in the text

Table 12 and 13 should be combined.Could you present these data on the plot together with SD and check if the values are statistically different or similar?

Table 16 and 17 should be combined.Could you present these data on the plot together with SD and check if the values are statistically different or similar?

 

Line 663: Why “ there must be a non-monotonic (in our case: quadratic) relationship 663 between yield fluctuations and dryness indicators”, some plots presented linear regression.

Author Response

Dear Reviewer 2, please find our responses in a separate file, attached.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors analyzed quadratic correlation between dryness indices and 2 wine grape yield to estimate future climate impacts in Hungary. The topic is interesting which will surply references for grape production in Hungary. The topic, data, methods and results are abundant and referrable. However, the authors emphasized some basic knowledges too much. The authors haven't focused  on the title topic well. Moreover, this paper is too big. I suggest make further revisions before acceptance. 

(1) Reorganize the manuscript. The different sections are too large, especially methods and Results. There are  18 figures and 17 tables. Too much. I suggest you put some similar figures and tables to Supplementary materials, and only briefly explain the supplementary figures and tables in the text. 

(2) Shorten the methods parts. Don't explain the same variable repeatedly (e. g, ETo or others). Also, number the equations please. Briefly introduce the methods if it is simple and common, when neccesary, cite the related references so that you don't need to explain the methods in redundant details.

(3)  Make a more fluent framework of manuscript. This is a paper, not a book. 

(4) The quadratic functional relationships seems have good performance in data fitting. However, since you only have five datapoints, you have artifically changed the trend of the datapoints. Sometimes the best fitting curves were not scientifically reasonable for interpreting the data characteristics.

(5) Too much references. Please removed some unnecessary citations, especially non-english or the old ones.

 

Author Response

Please, find our response to your comments enlosed!

Author Response File: Author Response.docx