Remotely Sensed Agroclimatic Classification and Zoning in Water-Limited Mediterranean Areas towards Sustainable Agriculture

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper “Remotely sensed Agroclimatic Classification and Zoning in water limited Mediterranean areas towards sustainable agriculture” uses a methodology to identify agroclimatic zones in three regions located at Spain, Tunisia and Lebanon. This methodology involves three sequential steps: i) a first classification using aridity and vegetation indexes, then ii) a zonification taking in account topographic and soil   features, and iii) the determination of suitability zones for annual crops using agrometeorological parameters like GDD, net radiation and precipitation.

The draft is in general well written with the need to improve/clarify some paragraphs. The organization of the manuscript is correct; however, in introduction L:52-71 the authors highlight the importance of climate change in the occurrence of extreme events and the influence that the spatiotemporal variability of temperature and precipitation has on the suitability and yields of crops, but in this work the effects of temporal variability, extreme events, or climate change on zoning are not considered, so I suggest removing this paragraph.

In materials and methods:

To delimit the WLGE zones, the authors use the VHI index calculated from the NDVI and LST values ​​obtained using Landsat 8 images from the period 2013 – 2020. They used for Spain, Lebanon and Tunisia, 640, 320 and 820 Landsat images, respectively. Several observations arise from this determination:

 

a) In the text it is not clear what procedure they use to obtain a single VHI value for each pixel and present the four vegetative drought classes of VHI in Figure 7. They need to clarify if they used mean, median, etc. values. And if the period of time in which these indices were calculated included the entire year or only the crop growth period (December - June).

 

b) The authors call this index (VHI) and the classes obtained from it as VHI “drought index”. Conceptually, I quote Jarraud (2006) verbatim. “Drought is an insidious natural hazard characterized by lower than expected or lower than normal precipitation that, when extended over a season or longer period of time, is insufficient to meet the demands of human activities and the environment. Drought is a temporary aberration, unlike aridity, which is a permanent feature of climate. Seasonal aridity, that is, a well-defined dry season, also needs to be distinguished from drought, as these terms are often confused or used interchangeably. The differences need to be understood and properly incorporated in drought monitoring and early warning systems and preparedness plans.”

Jarraud, M. (2006). Drought monitoring and early warning: Concepts, progress and future challenges. Concept paper of World Meteorological Organization. 26p.

In this paper the authors do not analyze the variability of the VHI around an average value, and they did calculate a single value from the average or median, so it would be an aridity index, as the authors demonstrate in the table 3, pairing the VHI with the AI. I suggest using “aridity” instead of “drought” throughout the text.

c) The authors do not explain why there is such a difference in the number of Landsat images used in each region. If it is due to cloudiness, they should carry out an analysis of the number of images available in each month in each region, given that if the lack of images coincides with the rainy season of the region (a topic that is not described in materials and methods) the classification could have a tendency towards more arid classes, given that little information would be available about the moment in which precipitation increases (and therefore increases in NDVI and decreases the LST). The number of images does not coincide with that later used to calculate the net radiation (L 582)

 

d) To complete the classification of WLGE zones, the authors use the IA index, calculated from precipitation and potential evapotranspiration obtained from the CHIRPS and MODIS satellite products (MOD16A2) with data from 2001 to 2020. Given the premise of the existence of climate change that the authors present in the introduction, these different periods cannot be compared, so in order to publish this work they would have to recalculate the AI ​​for the same period as VHI (2013-2020). The same applies for the calculation of GDD (2001-2020) and 20 days spring cumulative precipitation (1984-2020).

 

The authors must also clarify/justify the criteria of Ratings and suitability classes assigned to each criterion for agricultural suitability presented in table 4. For example, according to the DEM the poor category has a rating of 5, while slope poor class has a rating of 2 and landuse/landcover poor value correspond to 4 (other values ​​to justify are highlighted in the text). Why do WLGE zones have no rating values ​​and yet they intervene in the GIS multicriteria model?

 

It is also necessary to explain what the rules of the multicriteria model are, particularly how the rating values ​​were processed. And also clarify if a region is determined to be unsuitable by some criteria, it may be suitable in the final classification.

Figure 2 must be improved in order to understand it.

 

In table 4, four classes are presented for Landuse/Landcover, however in figure 10, the number of classes is higher and varies according to the study regions considered, the map classes must be unified.

 

Results

The figures that contain maps must unify the color scale considering the limits imposed in the classification presented in tables 1 to 4, (as in figures 7 and 8), however in figure 10 the DEM is presented in meters with different Color scales for each of the study regions and the slope map is not presented. Something similar occurs in Fig. 12 where the GDD values ​​for wheat are presented with a different color scale for each region without highlighting the minimum value necessary for the growth and development of wheat. Similar considerations should be taken into account for figures 13 and 14. In maps that share a legend, this should be presented only once and without covering the study area (see figures 7,8,9,10 (only DEM, Slope and Land use/Landcover), 12, 13, 14, 15, 16. The maps must be presented in a similar way (all with a mask of the study region or not. See figs 10b LULC, 15b and 16b)

 The analysis is only carried out for wheat, so I suggest changing the title and objectives.

 Minor comments are highlighted in the draft.

Comments for author File: Comments.pdf

Author Response

Dear reviewer we would like to thank for your valuable comments and recommendations. We responded to your comments one by one.

Please see the attachment

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript has strong research significance, and the whole manuscript organized well, and achieved relatively good validation analysis results. I suggest accept it to publish in the journal. 

However, there are still some problems in this manuscript that need to be further answered or improved, as follows:

1. WLGE is created using a combination of VHI and AI, which is divided into five regions. Since climate drought leads to agriculture drought，why cannot VHI be used directly?

2. How to verify the category indicators in Table 4 are correct? How to verify the reliability of agroclimatic classification results?

Author Response

Dear reviewer we would like to thank for your valuable comments and recommendations. We responded to your comments one by one.

Please see the attachment

 

 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Only few highlighted and commented typing modifications remain in the draft

Comments for author File: Comments.pdf

Author Response

Dear reviewer,

II have implemented the corrections you suggested on pages 11 and 22. The changes have been highlighted in turquoise. Please see the attachments

Author Response File: Author Response.docx