Drought Characterization and Trend Detection Using the Reconnaissance Drought Index for Setsoto Municipality of the Free State Province of South Africa and the Impact on Maize Yield

Round 1

Reviewer 1 Report

       This study evaluated the drought characteristics in the Setsoto district of South Africa from 1985-2019 using the RDI. Yearly and spatial variations of RDI was investigated, and yield maize loss was correlated with RDI as well. This revised manuscript has considerable improvements compared with the previous one. But there are still some concerns that should be addressed before publication. Please refer to the following comments for the details.

 

1. L22: This statement about El Nino is quite questionable.
2. L98: Since the aim of this study is to connect drought with maize, please add more details of crop phenology, water-stress coefficient, and distribution.
3. L198: Figure 1 should be improved.
4. L260: The accuracy of PET calculation is suggested to be checked.
5. L367: One previous comment: The deviation of drought condition and yield were opposite in some years. Please explain the possible reason.
6. L471: Sub-section titles are suggested to be added in this section.
7. A lot of grammatical mistakes should be corrected throughout the text.

Author Response

 

Dear Reviewer,

Thank you for the contributions and suggestions which have added value to the manuscript. The entire manuscript has been edited by a professional editor whose home language is English. We have carefully attended to each of the comments and suggestions and can be seen in the documents submitted to the editor. 

Reviewer 1: L22: This statement about El Nino is quite questionable.

Authors:  Some additional analyses were done to assess the correlation between El Nino years and the RDI values but there was only one year in which it was significant and data were not available for long enough to establish a more robust relationship so the statement was deleted

Reviewer 1: L98: Since the aim of this study is to connect drought with maize, please add more details of crop phenology, water-stress coefficient, and distribution

Authors: Additional text has been added to place in context the time of planting and the time of harvest of the maize crop and the various stages in the crop phenology. The crop is at its most vulnerable, to water stress during germination and establishment, silking, pollination and grain filling and this is covered by the RDI 3 - and 6-month analyses. Information is included in the text on crop distribution and the water-stress coefficient is not included explicitly because PET was used together with rainfall to calculate the RDIs. 

Reviewer 1: 1  L198: Figure 1 should be improved.

Authors: An improved map has been included

Reviewer 1: L260, the accuracy of PET calculation is suggested to be checked.

Authors: The equation has been corrected by including the correct superscript

Reviewer 1: L367: One previous comment: The deviation of drought condition and yield were opposite in some years. Please explain the possible reason.

Authors: This is very difficult to explain because of a number of confounding factors, including the exact timing of a drought event, the magnitude of that event, the duration of the event, together with soil water status and PET. Text to this effect is included in the manuscript and even though it cannot be fully explained it remains an interesting speculative finding.

Reviewer 1: L471: Sub-section titles are suggested to be added in this section.

Authors: Subsection titles have been added to the manuscript

Reviewer 1:  A lot of grammatical mistakes should be corrected throughout the text.

Authors: The document has been professionally edited, hopefully eliminating all grammatical errors.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors did lot of work during revisions and now manuscript looks in much better shape. There are some minors concerns which need some attention.

1. Scientific name of each crop should be added at its 1st appearance, for instance maize at line 71. Check whole manuscript critically.
2. Authors use tones/hectare and tons ha^-1 terms. Need to be uniform and use tons ha^-1 throughout the manuscript.
3. Conclusions section should be reduced up to 50% of its current volume.

Moreover the whole manuscript should be read carefully to remove spell and language mistakes.

Author Response

Dear Reviewer,

 

Thank you for the contributions and suggestions which have added value to the manuscript. The entire manuscript has been edited by a professional editor whose home language is English. We have carefully attended to each of the comments and suggestions and can be seen in the documents submitted to the editor. 

Author Response File: Author Response.pdf

Reviewer 3 Report

Drought Characterization and Trend Detection using the Reconnaissance Drought Index (RDI) for Setsoto Municipality of the Free State Province of South Africa and the Impact on Maize Yield

By HB Abubakar et al.

Submitted for publication : Water

This manuscript has been substantially improved. However it remains difficult to read, its progression is tortuous and we still find many weaknesses. I think improvements are required to make the manuscript publishable.
The authors chose to characterize the drought using RDI . Their presentation of RDI is very partial and insufficient. See detailed description in Tigkas et al. 2015 and reference to the software they used. A reference that must be cited (Tigkas, D., Vangelis, H., & Tsakiris, G. (2015). DrinC: a software for drought analysis based on drought indices. Earth Science Informatics, 8(3), 697-709). . I'm not opposed to RDI but at the current time SPEI is very unanimously used in our communities. As a result, it allows us to have an unified vision of drought at all scales from the plot to the globe.
For RDI they use 3 time windows October to December, October to March and finally October to September. It is necessary that these choices are argued according to the phenological calendar of maize since the authors are looking for a link between drought and production.
In these three cases the calculation of rank correlations between RDIs and yield would give an objective answer to the question: Which RDI is best able to account for a significant part of the variability of production?

The link between your RDIs and El Nino you report are not documented. This can be easily corrected by downloading the Oceanic Nino Indices data from the NOAA website. http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/

Here too rank correlations would be relevant. This contribution would make it possible to link this issue with the investigation of the trend demonstrating an incresing dryness that the authors propose. Indeed, the frequency of intense El Nino episodes seems to increase under the effect of global changes.

My last comment concerns the mapping of RDIs. The maps are based on only 4 sites distributed over a very large territory. The sources of variations in monthly rainfall amounts or temperatures and thus of PET are not specified. The choice of the inverse distance weighting interpolation method should be justified. I don't believe that RDIs that are derived from complex non-linear transformations can be so easily interpolated. A trial by first interpolating rain and PET and finally deriving RDI maps would be necessary.
Minor comments

Please it will be useful to give PET values for your sites and perhap to compare your estimates with litterature values from Penman Monteith equation applied in your area.

The map in figure 1 must be improved

Table 2 Are you sure about your CV?

Line 447 «The spatial display maps for RDI values were prepared using Arc GIS 10.3» please move to MM

Author Response

Dear Reviewer,

 

Thank you for the contributions and suggestions which have added value to the manuscript. The entire manuscript has been edited by a professional editor whose home language is English. We have carefully attended to each of the comments and suggestions and can be seen in the documents submitted to the editor. 

Thank you

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

       The authors have positively addressed all my concerns. Though the adopted method is simple, the drought analysis and results in the Free State Province of South Africa are worth publication. I recommend its publication while I also think there are a lot of additional drought analyses should be conducted in South Africa from now on. In particular, when considering the Cape Town had nearly run out of its water during these years, I encourage authors investigate more special drought features for this special land.

Author Response

Reviewer 1:  I recommend its publication while I also think there are a lot of additional drought analyses should be conducted in South Africa from now on. In particular, when considering the Cape Town had nearly run out of its water during these years, I encourage authors to investigate more special drought features for this special land.

Authors: Thank you for your contributions and suggestions, we appreciate that as the climate is getting drier in South Africa, the scientific community will need to conduct more analyses that can be used in early warning systems to inform farmers and minimize the risk associated with agricultural production.

Author Response File: Author Response.docx

Reviewer 3 Report

In my last minor comment I would like you introduce your GIS application and interpolation methodology in the section dedicated to material and methods.

Even a very rough comparison of your PET estimates on an annual scale and on one station with Penman-Monteith would be a plus in the validation of your results.

Author Response

Reviewer 2: In my last minor comment I would like you introduce your GIS application and interpolation methodology in the section dedicated to material and methods.

Authors: Further information has been added to the text to the effect.  Geographical Information System (GIS) techniques are important for agricultural modelling in  spatial and temporal analysis of variables such as rainfall, soil and temperature[1, 2]. To understand the spatiotemporal pattern and occurrence of drought in the study area GIS interpolation techniques was applied. In this study, we used the inverse distance weighted (IDW) method of the GIS spatial-interpolation technique for interpolation of the RDI data

 

Reviewer 2: Even a very rough comparison of your PET estimates on an annual scale and on one station with Penman-Monteith would be a plus in the validation of your results.

Authors: Thank you so much for your suggestion. We agree with the suggestion but in order to calculate Penman-Monteith, one needs good data ont radiation. There are only very few weather stations in South Africa, especially those in urban areas, which have these data records. The data are not useful for the agricultural areas in which we are working. We don’t have data for the crop net radiation. The PET was calculated via the “Drinc” software which allows only the three methods namely (Hargreaves, Thornthwaite and Blaney- Criddle).

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

General Comment

The manuscript written by Abubakar et al. analyses the drought tendencies of a water-scarce region in South Africa. It contributes to the understanding of the hydrological cycle in susceptible environments linked to agriculture. The manuscript is worth of publication in Water after a series of major changes required to improve the quality of the scientific analysis.

Major Comments

1. The manuscript's aim is completely missing from the manuscript. It is necessary to state it in the manuscript, preferably at the end of the introduction as well as in the abstract. Without it, is not possible to assess objectively the manuscript outcomes.

 

2. The authors should contextualize the importance of analyzing the drought patterns in this specific area of South Africa. This information is completely missing in the introduction. Some questions that will help the authors to improve are:

               - Why Setsoto Municipality?

               - What is the importance of Maize in the region?

               - What is the relationship between crop management and drought patterns (lines 44-49)?

               - What is the actual knowledge of drought patterns for South Africa?

               - What indices have been used in South Africa or the target region?

These questions should help the authors to improve the introduction. It is necessary to provide a better context for the manuscript.

 

3. The methodology should be more detailed. From the map is possible to know the relative location of the study site. However, if the authors include more information (e.g., elevation map, land cover map, rivers, etc.) this will help to provide a better analysis linking other factors such as the topography as a possible source of influence on the drought patterns/effects. It is not possible to relate the size of the research area in terms of km² (e.g., the scale on the map is not linked to any of the images).

 

4. Map. It should be improved.

 

5. There is no information on the yield data source. Also, it is not possible to link the production area to different stations (influence area). It is necessary to add more information such as:

               - Number of farmers, average size (ha)

               - Production area

               - Type of crop (rain fed or irrigated)

 

6. The authors should be careful with the description of parameters and variables. Also, with the description of the equations and formulas used in the manuscript. Some examples are:

               - The term potential evapotranspiration is expressed as PET, this can be interpreted as Precipitation * Evaporation * Temperature. This term should be expressed as E_pot.

               - The expression for the index RDI should be I_RD

               - The I_rd index should be described better. The authors should follow a similar description as https://doi.org/10.1016/j.jaridenv.2012.07.020

- How E_pot. Is calculated? The authors should provide formula and assumptions.

 

7. The authors can describe better the relationship between yield and I_RD with scatter plots. In this way, it is possible to link drought conditions with low/high yields. 

 

8. The results section misses completely the historical pattern of precipitation, temperature, and maize yield for the different stations and regions. This information is important to understand the drought patterns and the effect on the yield. Also, table 2 shows the mean values for the yields, however, the standard deviation shows a large coefficient of variation. This is the reason why historical patterns are needed.

 

Before these improvements, the discussion should be improved and include the impact of drought on the maize yield. This is completely missing from the manuscript and is the richest part of the analysis. Also, conclusions should be updated.

 

 

Reviewer 2 Report

       This study investigated the drought events and trends for Setsoto Municipality of South Africa from 1985-2019 using Reconnaissance Drought Index (RDI) and MK test with Sen’s slope. In particular, precipitation and temperature data from four weather stations were used, and the implication on maize yield had been roughly analyzed. Based on the obtained results, the authors tried to find out the historical drought characteristics and provide some suggestions for agricultural production.

First of all, this work is within the scope of WATER and may has considerable impacts on the application domain. But some important issues have to be solved before considering publication. My comments have been provided below for the authors when improving it further.

• L48-49: This sentence does not make sense to me, please double check it.
• L62-L63: The motivation of this study is unclear and unreliable. There are plenty of drought studies both at global and local scales. I suggest a more careful review should be conducted. And more importantly, the significance of local study in South Africa should be explicitly provided, which can help us to understand the importance of this work.
• L87: Please add the latitude and longitude information in this figure to help us know exactly the location of this study area.
• L90: Is there any pre-processing of these meteorological data? Because it is common that the meteorological data is temporally discontinuous, or there are some outliers.
• L100: Since the locations of these four stations have been marked in the Figure 1, we do not need this table any more.
• L107: The last word ‘and’ in this line is unnecessary. Actually, there are so many typos in this manuscript that I strongly suggest a careful review of this manuscript.
• L110: The reason for choosing RDI, instead of SPEI or VegDRI, should be provided here.
• L158: Based on the results of Figure 2-4, the authors found ‘the yield declined in that year or in the following year’. So please discuss that why the yield in following years can be influenced by the previous year. We should not only show the obtained results. The soundness and scientificness behind these results should also be analyzed.
• L162: The deviation of drought condition and yield were opposite in some years. Please explain the possible reason.
• L188: The authors stated that ‘but moderate droughts had little influence on yield’. Please also analyze this point further.
• L213: The section title of ‘3.7’ should be deleted.
• L221: There is no spatial analysis about drought distribution in this study area. This is also another serious deficiency of this study.
• L242: The ‘strong El Nino event’ has little discussion or supporting information in this study. The same sentence has also existed in the abstract.
• L273: The word ‘c’ should be deleted.
• L286: The conclusive sentence that ‘the first 3 months of the growing season are critical for the .. yield’ should be elaborated further.
• The academic contribution of this study to the drought community is unclear. The innovation of this study is also missing. In the revised manuscript, if there is a chance for revision, please enhance this point to demonstrate the solid reason for publication.