Seasonal Mapping of Irrigated Winter Wheat Traits in Argentina with a Hybrid Retrieval Workflow Using Sentinel-2 Imagery
Round 1
Reviewer 1 Report
The article presents interesting outcomes of wheat traits retrieval based on sentinel 2 data and hybrid methods. It is a well written paper, with comprehensive methodology, with clear structure.
General comment: Although the paper is well organized, it is quite long for a research article. Although the journal has no page limits, the authors can reduce the volume of the paper, as some information does not add significant information (see specific comments below).
Line 68 delete which
Line 106: However, the suitability of hybrid retrieval methods for irrigated winter wheat monitoring over Argentinean study sites remains yet to be investigated.
Figure 1 and 2 can be combined in one panel figure.
The line gives an impression that the main gap that the research is filling is tied to the crop and study area. This is not a very strong argument for the novelty of the paper. I would recommend here to put the focus on other novel aspects (i.e. active learning, estimation of not only LAI, but also other traits).
Line 244 -248: As the paper focuses on winter wheat, I would suggest skipping the description of all other crops grown in the area ( as it does not affect later the results and discussion).
I would recommend replacing Agri-environments with cropping systems.
Section 4.4 Limitations of the study: maybe the authors can mention that the study was conducted with data from one growing season, so temporal transferability is still an open question.
Line 698 The here >> delete the
Line 698-702 This section of the conclusions is redundant, as it repeats the information from the introduction/method section and potentially can be deleted.
Based on this I would recommend accepting the paper with minor revisions.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The manuscript aims to using a hybrid approach based on PROSAIL and GPR-AL to generate temporal and spatial maps of LAI, CCC and VWC in wheat crop from multi-temporal sentinel-2 satellite data. The manuscript is well written, and the story is easy to follow. The manuscript may be considered by the journal after a minor revision. My comments mainly include,
(1) The novelty of the current study compared to the previous studies is suggested to be clearer.
(2) In general, sample size is larger than spatial resolution of satellite data for reduced uncertainties caused by field data collection and preprocessing of satellite data. However, the authors did not consider this their measurements and study. Explanations on this should be provided.
(3) The authors may reduce the space to introduce the GPR in the Section 2.2. Detailed information may be present in the appendix document.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
This paper examined a hybrid method to retrieve LAI, CCC, and VWC for winter wheat, which is an elaborate and reasonable approach. This paper can be considered publication after a bit of modification about the application of machine learning.
The authors can compare the result of GPR with other AI methods, such as random forest (RF) and deep neural networks (DNN). Then, the authors can show the advantages of GPR and the hybrid model more effectively.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf