Estimating Farm Wheat Yields from NDVI and Meteorological Data

Round 1

Reviewer 1 Report

The paper “Estimating farm wheat yields from NDVI and meteorological 2 data” focuses on the evaluation of the NDVI integral, peak NDVI, monthly precipitation, monthly temperature for setting up an empirical winter wheat yield model for northern Belgium. For achieving this evaluation, the authors have analyzed two separate random forest models for establishing which NDVI (a NDVI or maxNDVI) better predicts winter wheat yield. Data retrieved from estimations were analyzed by comparison with weather conditions from fields (Tmin, Tmax and P). The Person correlations revealed small correlation between NDVI and climate variables. Winter wheat yield variability was better predicted by monthly precipitation during tillering and anthesis than by NDVI (R²=0.66). In conclusion, winter wheat yield modelling using NDVI derived yield proxies as predictor variables is dependent on the environment.

The authors applied the researches from reference [4] without to consider the soil effects and topography of the aria. Is there any soil degradation in the area? What is the novelty of this article?

Author Response

Dear reviewer,

We appreciate the interesting questions and remarks on our manuscript. Below we will respond to your specific questions.

• Concerning your question on the effect of soil and topography in the area:

Information on soil was indeed not included directly in the model. This was also not the case in the referenced research [4, 6]. However, in the discussion section we mention that adding information on the location of the fields (i.e. in which of the seven agricultural regions of northern Belgium the field is located) did not improve the model (lines 273 and 276). Since, each agricultural region is characterized by a distinctive mixture of soil and terrain conditions the effect of soil and topography of the studied fields on the model performance was tested indirectly. In a similar way the location of the fields were used in the yield modeling in the referenced research [4] (see lines 266-273). A similar approach was developed by [6] in northern France, where the combined use of NDVI and meteorological variables performed well for different soils and management regimes without incorporating these variables directly into the models. The agricultural regions of northern Belgium were added to Figure 1.

• Concerning your question on the novelty of this article:

In this research we demonstrated that in order to use NDVI as predictor in empirical crop yield models, the NDVI series have to be sensitive to yield affecting weather conditions during important phenological stages such as tillering and anthesis . To our knowledge this was not shown before for NDVI based wheat yield models. The environment dependency of wheat yield models based on fAPAR, another vegetation index, was already demonstrated by [23] as mentioned in the discussion.

All the best,

The authors

Reviewer 2 Report

The authors found that the NDVI series were weak predictors of winter wheat yield in northern Belgium, and winter wheat yield variability was better predicted by monthly precipitation during tillering and anthesis than by NDVI derived yield proxies in the period from 2016 to 2018. The conclusion was provided that winter wheat yield modelling using NDVI derived yield proxies as predictor variables is dependent on the environment. This information is valuable for farmers and decision makers.

Author Response

Dear reviewer,

We appreciate your remarks on our manuscript. The revised manuscript can be found in attachment.

All the best,

The authors

Reviewer 3 Report

Comments to the Authors

The entitled manuscript of Estimating farm wheat yields from NDVI and meteorological data is organized well and providing valuable information about setting up an empirical winter wheat yield model for northern Belgium.

In the introduction section, I did not observe that the authors provided sufficient information about the importance of winter wheat in northern Belgium. They should cover that gap and write at least a paragraph that provides significant information about winter wheat at that region.    

Specific comments:

Line 16: provide the scientific name for wheat, also where it is mentioned first in the introduction section.  

Line 71: re-write this sentence “Winter wheat yield from 2016 to 2018 in northern Belgium was studied (Figure 1).”

Line 72 : remove this statement “Winter wheat is an important crop in northern Belgium.” from M&M to the introduction section. Also, the authors should provide a paragraph in the introduction explaining the importance of winter wheat in northern Belgium along with some factual information that would enable the reader to understand the reader how this given crop is important to the farmers and eventually to Belgium.

Line 143: revise to May [DOY(day of year) 151], apply it where you have a similar thing.

Line 308: revise “When we look at NDVI values in the months January, February and June we did not see a negative effect of high precipitation values on NDVI (Figure 8)” …. to …..when we examine NDVI values in the months January, February and June we did not observed see a negative effect of high precipitation values on NDVI (Figure 8). Please, utilize academic words in your manuscript and make sure that you will check the entire manuscript and revise where it needs revision.

Author Response

Dear reviewer,

We appreciate the interesting questions and remarks on our manuscript. We revised the manuscript with care and provide a response to your remarks below.

All the best,

The authors

• cover that gap and write at least a paragraph that provides significant information about winter wheat at that region.

We added some sentences on the importance of winter wheat in northern Belgium. Lines 64-67

Specific comments:

• Line 16: provide the scientific name for wheat, also where it is mentioned first in the introduction section.  

This was added.

• Line 71: re-write this sentence “Winter wheat yield from 2016 to 2018 in northern Belgium was studied (Figure 1).”

We decided to delete this sentence since this information was covered at the end of the introduction.

• Line 72 : remove this statement “Winter wheat is an important crop in northern Belgium.” from M&M to the introduction section. Also, the authors should provide a paragraph in the introduction explaining the importance of winter wheat in northern Belgium along with some factual information that would enable the reader to understand the reader how this given crop is important to the farmers and eventually to Belgium.

This was moved to the introduction section see lines 64-67

• Line 143: revise to May [DOY(day of year) 151], apply it where you have a similar thing.

This was changed. Lines 145, 155-158 and 282

• Line 308: revise “When we look at NDVI values in the months January, February and June we did not see a negative effect of high precipitation values on NDVI (Figure 8)” …. to …..when we examine NDVI values in the months January, February and June we did not observed see a negative effect of high precipitation values on NDVI (Figure 8). Please, utilize academic words in your manuscript and make sure that you will check the entire manuscript and revise where it needs revision.

This was changed and revised for the entire manuscript.

Round 2

Reviewer 1 Report

Dear Authors,

Thank you for reply. 

Concerning your comments on the effect of soil and topography in the area, since sandy soils are dominant your results confirm the real situation: without irrigation the yield production of the Triticum aestivum will be always lower. As you know Triticum aestivum prefers moist soil.

Empirical models for yield estimation consider NDVI correlated with meteorological data and soil data. In this situation, what is the novelty of the paper compared to [4]?