A Hybrid Prediction Model for CatBoost Tomato Transpiration Rate Based on Feature Extraction
, Xiaolong Zhang 4, Baijuan Wang 5,* and Wengang Zheng 3,*
Round 1
Reviewer 1 Report
I have recorded my comments on various pages of the attached PDF, which are self-explanatory.
Comments for author File: 
Comments.pdf
Author Response
Reviewer #1:
(Comment 1) Write few sentences about the relationship/importance of evapotranspiration and irrigation requirement for tomato crop giving references. (Introduction)
Reply: Thank you very much for your review comments. We have added the relevant content in lines 42-48 as per your suggestion. The specific additions are as follows:" The transpiration intensity of tomato is closely related to the amount of irrigation, and many studies have found that ETc decreases with decreasing irrigation [2]. Therefore, reasonable irrigation regulation according to changes in the greenhouse environment and crop growth stage is an important means to ensure the growth of greenhouse crops [3]. Among them, accurate and quick prediction of transpiration rate (Tr) of greenhouse crops is the key to make irrigation regulation reasonable. "
(Comment 2) Light intensity should be in umol/m2/s1 (micro mole) or in MJ/ m2/s1. You can convert lux to these ones (PPFD).
Reply: Thank you very much for your comment, we have changed the lux units to umol/m2/s1.
The revised table is shown below:
(Comment 3) What is e? Mention here. (formula 2)
Reply: Thank you for the correction to the formula, the letter "e" stands for natural constant. We have added the description of the letter e in the formula on line 145.
(Comment 4) Please give few references here to support your reasoning/assumptions. (Line 413-419)
Reply: Thank you for your suggestions on this paper, we have added relevant literature as per your suggestions, e.g., a total of three papers from 27-29.
Author Response File: Author Response.pdf
Reviewer 2 Report
The manuscript provides up-to-date knowledge how to improve the prediction accuracy of models by the extraction of the data characteristics during the vegetation period of tomato grown under greenhouse conditions. The development
The increasing demand for understanding how the environmental factors determine plant growth at different phenophases justifies the actuality of the study (even this is not highlighted in the manuscript clearly.
The manuscript is based on an intensive study with high scientific soundness, especially in terms of modelling, but the practical justification of the study is also mentioned (can be improved in the conclusions section).
The article highlights some (not too many) national and international references relevant to the topic.
The study is quite complete, though I feel strong dominance of the modelling part at the expense of the experimental part. Maybe the title could be modified accordingly, and also some technical terms should be determined more clearly.
Among the results, some are too obvious, they would not need such an intensive study to be determined. I suggest to put more emphasis on the modelling results.
The tables and figures are self-explanatory and well edited.
The discussion of the results is well structured and sufficiently detailed citing relevant references.
The Conclusion part contains statements based on own results. Nevertheless, I suggest to extend it with another sentence mentioning the concrete possibility of the interpretation of the results for the practice.
I inserted several sticky notes with my specific comments, questions and corrections in the pdf file of the manuscript.
Comments for author File: Comments.pdf
I recommend to revise some sentences in terms of English (especially the use of past tense).
Author Response
Reviewer #2:
(Comment 1) The title should be revised expressing more the dominance of modelling.
Reply: Thank you very much for your comment. We have completed the modifications to the title, with the latest modifications as follows: "A Hybrid Prediction Model for CatBoost Tomato Translation Rate Based on Feature Extraction", which emphasizes modeling.
(Comment 2) Even it is mentioned in the abstract, use full name of the abbreviations when mentioned first time in the text.
Reply: Thank you very much for your comment, we have completed the abbreviations mentioned for the first time in the abstract and the text as follow: “transpiration rate (Tr), Competitive adaptive reweighted sampling (CARS), t-Distributed Stochastic Neighbor Embedding (t-SNE), root mean squared error (RMSE), etc.”
(Comment 3) Use past tense to describe a terminated experiment.
Reply: Your comments are very much appreciated. We have addressed the issue of incorrect use of word tenses.
(Comment 4) Actually this is a weighing lysimeter. (Line 117-118)
Reply: Thank you for your valuable suggestion. We have modified the "Intelligent Online Matrix Weighing System" to "weighing lysimeter" according to your suggestion.
(Comment 5) This is confusing (seems like anyzhing to do with rain). Actually this is the leachate (or drain water). (equation (1))
Reply: Your comments are very much appreciated. We have corrected Drain to Dliquid.
(Comment 6) These results are obvious involving nothing novel, too obvious! (Line 301-311)
Reply: Your comments are very much appreciated. We are very sorry for the error in our previous analysis that led to the error within the graph, we have corrected the image as shown.
(Comment 7) Revise this part. (Line 317-318)
Reply: Thank you for your comments. We have revised the original sentence to read: "The number of environmental variables shows a gradual decreasing trend, and tends to stabilize as the number of sampling times increases."
(Comment 8) these transpiration values could easily be expressed in more convertible mm/h and mm/day units (1000 g/m2 = 1 mm) (Table 4.)
Reply: Thank you for your comments. We have made changes based on your suggestion. The revised table is shown below.
(Comment 9) This paragraph is rather a summary than conclusions – revise. (Line 473-475)
Reply: Thank you for your comments. We modified the first paragraph of the conclusion to read as follows, "In this study, we analyzed and extracted the main environmental variables affecting tomato transpiration, and established a hybrid prediction model for tomato Tr based on CatBoost algorithm (CARS-CatBoost model). "
(Comment 10) How? Add some concrete details as that would be the practical significance and utilization of the study.
Reply: Thank you for your comments. The content of this question has been added in lines 459-467 of the article.
Author Response File: Author Response.pdf
Reviewer 3 Report
The review of the literature should include numerous available works on the irrigation of tomatoes in greenhouse cultivation and the transpiration of tomatoes. Empirical techniques and devices for direct measurement of Tr should also be mentioned. Why do the authors write about 6 parameters measured in the experiment, and mention only three of them (lines 18-19)? Since the experiment was conducted in 2022, it should be written in the past tense, not the present (is/was). The sentences "The irrigation method is to control irrigation based on the radiation accumulation." (lines 105-106) needs to be changed to make sense (irrigation control). The term "crop Tr is used as the basis for irrigation" (lines 108-109) is insufficient to describe irrigation control techniques and needs to be supplemented. The use of the WTS-300 device (line 116) in research requires supplementing the article with other publications (research/experiments) in which it was used (validation). How did the described 10-minute interval of measuring environmental parameters translate into determining the times and doses of irrigation (line 137)? What were the reasons for recording incorrect data (or missing data) described by the authors, how many (%) were the cases in relation to the total data (lines 148-150)? It would be interesting to compare the measured temperatures Tmax/max, Tmin/min presented over time. What caused such large differences between the max and min temperatures measured in the conditions and time of the experiment? Why do the authors write about recording data once every hour (line 456) and once every 10 minutes (line 137)? In Conclusions, the authors write about the application of the obtained results to control irrigation. Such a statement would have to be proved by providing a detailed description of irrigation control during the experiment.
Author Response
(Comment 1) The review of the literature should include numerous available works on the irrigation of tomatoes in greenhouse cultivation and the transpiration of tomatoes. Empirical techniques and devices for direct measurement of Tr should also be mentioned.
Reply: Thank you for your careful review of this manuscript. We have added the relevant content in lines 42-48 as per your suggestion. The specific additions are as follows:" The transpiration intensity of tomato is closely related to the amount of irrigation, and many studies have found that ETc decreases with decreasing irrigation [2]. Therefore, reasonable irrigation regulation according to changes in the greenhouse environment and crop growth stage is an important means to ensure the growth of greenhouse crops [3]. Among them, accurate and quick prediction of transpiration rate (Tr) of greenhouse crops is the key to make irrigation regulation reasonable. "
(Comment 2) Why do the authors write about 6 parameters measured in the experiment, and mention only three of them (lines 18-19)?
Reply: Thank you very much for your careful consideration of the structure of this paper, your suggestions are very valuable to improve this paper. We have already made the parameter additions in lines 22-23 of the article.
(Comment 3) Since the experiment was conducted in 2022, it should be written in the past tense, not the present (is/was).
Reply: Thank you very much for your careful reading of this manuscript, and we have corrected the tense problem you pointed out.
(Comment 4) The sentences "The irrigation method is to control irrigation based on the radiation accumulation." (lines 105-106) needs to be changed to make sense (irrigation control). The term "crop Tr is used as the basis for irrigation" (lines 108-109) is insufficient to describe irrigation control techniques and needs to be supplemented.
Reply: Thank you very much for your suggestion, we have made the changes you suggested and lines 111-115 read as follows: "Irrigation is controlled by setting a radiation accumulation threshold. The radiation accumulation threshold set for each irrigation is 120 Kj·m-2·h-1, and when the threshold is reached the irrigation controller activates the water pump and solenoid valve to start irrigation."
(Comment 5) The use of the WTS-300 device (line 116) in research requires supplementing the article with other publications (research/experiments) in which it was used (validation).
Reply: Thank you very much for your comment, this equipment was designed and developed by the National Agricultural Intelligent Equipment Engineering and Technology Research Centre and has relevant patents, to which we have added relevant literature as proof.
(Comment 6) How did the described 10-minute interval of measuring environmental parameters translate into determining the times and doses of irrigation (line 137)? What were the reasons for recording incorrect data (or missing data) described by the authors, how many (%) were the cases in relation to the total data (lines 148-150)?
Reply: Thank you for your review comments on this study. We have explained and added to the issues you mentioned in lines 156-159 of the text as follows:" In the process of monitoring and transmitting data using IoT sensors, issues such as device stability and signal quality have resulted in partial data loss, duplication, data imbalance, and inconsistent data types. The number of abnormal data accounts for 1.7% of the total dataset."
(Comment 7) It would be interesting to compare the measured temperatures Tmax/max, Tmin/min presented over time. What caused such large differences between the max and min temperatures measured in the conditions and time of the experiment?
Reply: Thank you for your careful reading of this manuscript, The following is an explanation of the problem: Temperature and relative humidity are collected at 10-min intervals, and the temperature and relative humidity values during the hour are counted in terms of maximum, minimum and average. For example, 13:20 noon on the sunny day of 2 July has the highest temperature (43.5°C) in that hour, which is also the highest temperature of Tmax in the measurement period, and 12:00 midnight on 30 May has the highest temperature (12.4°C) in that hour, which is also the lowest temperature of Tmax in the measurement period, and the range of 12.4°C-43.5°C constitutes the Tmax in the measurement period. maximum, minimum and average temperatures; conversely, 13:50 noon on a sunny day on 2 July has the minimum temperature (41.8°C) for that hour, 12:50 pm on the night of 30 May has the minimum temperature (12.4°C) for that hour, and 12.4°C-41.8°C constitutes the maximum, minimum and average temperatures for Tmin during the measurement period, and so on. I hope this explanation satisfies you.
(Comment 8) Why do the authors write about recording data once every hour (line 456) and once every 10 minutes (line 137)?
Reply: Thank you very much for your review comments. We set the acquisition frequency of the sensors to 10min/time, this is to reduce the possibility of data loss due to equipment problems, and finally the last data acquired during each hour was analysed as the data for that hour. We hope you will be satisfied with our session.
(Comment 9) In Conclusions, the authors write about the application of the obtained results to control irrigation. Such a statement would have to be proved by providing a detailed description of irrigation control during the experiment.
Reply: Thank you for your comments. We've made additions to the article on lines 461-469 that we hope will satisfy you.
Author Response File: Author Response.pdf
