Identifying Dry-Season Rice-Planting Patterns in Bangladesh Using the Landsat Archive
Round 1
Reviewer 1 Report
This work aims to identify spatio-temporal patterns in dry season rice production for all of Bangladesh in order to better understand the current status and potential of double-cropped rice systems, as well as to provide a more robust platform for rice system monitoring in areas without in situ data. Specifically, it integrates spatial and temporal aspects of rice classification into a procedure which is regionally flexible and requires few parameters as model input. It constructs a time series of Landsat data based on a harmonic model to overcome the sparse availability of observations (~16 days) and minimize the effect of different noises. Therefore, the objectives of this work were: (1) To demonstrate the use of an HTS-VIs model with Landsat in providing consistent estimates of areas annually planted to boro rice in Bangladesh from 2013–2018; (2) To analyze regional differences in boro rice phenology and production; and (3) To compare and contrast the results of the HTS-VIs model with reference sources for boro rice areas in Bangladesh. The paper is clearly presented and is supported by the data and the techniques used. The results seem to be reasonable and the method seems to be consistent. Since the vegetation indices are the main information to develop this method, it is necessary to make clear if it is using EVI or EVI2 (since they have different formulas) in this study. It seems that it is using EVI as observed through the text.
Some comments and suggestions:
L. 62 The Moderate Resolution Imaging Spectro-radiometer (MODIS) constellation --- The Moderate Resolution Imaging Spectro-radiometer (MODIS) sensors
L. 70 its coarse resolution --- moderate spatial resolution
L. 170 followed by water infiltration and flooding, which slowly transitions to light --- which slowly makes transition to light
L. 188 Enhanced Vegetation Index (EVI2) [43–45], --- If it is using EVI, then the references should be changed.
L. 191 that EVI2 (referred as EVI from now on) and --- If it is using EVI, this sentence should be removed
L. 197 EVI formula and not a EVI2 formula ? --- However it is using EVI2, the formula should be changed
L. 215 LE7 and LC8 satellite platforms image(d) the entire earth --- imaged
L. 265 ensure that EVI is has a positive ---- ensure that EVI has a positive
L. 301 with a coarser spatial resolution but with more frequent return time (1 day). --- a moderate spatial resolution (related to MODIS sensor)
L. 439 offered by the constellation. --- offered by the satellites (or platforms)
Author Response
Dear Reviewer,
Please see our point-by-point response attached in the pdf. The authors greatly appreciate the efforts of the reviewer to improve our manuscript. We believe your critiques and suggestions, along with our responses, have enhanced this research.
Thank you.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear Authors
I revised the manuscript “Identifying dry season rice planting patterns in Bangladesh using the Landsat Archive” submitted to the Remote Sensing Journal. The paper is very interesting and used GEE and Landsat archive to identify characteristics of rice phenology in Bangladesh but I have some concerns which need to be addressed before considering for final publication.
Overall comments:
- Please discuss more about the ruleset for rice pixel identification with relevant references. Is there any previous model, which used the similar ruleset?
- Please discuss more about the fieldwork or ground truthing of your data.
Minor comments:
- Line 66, please use Synthtic Aperture Radar before SAR.
- Line 71 and 104, the size of average rice fields in Bangladesh has repeated. Please discard the second one in line 104.
- Line 105, please use vegetation indices before VIs, since it is the first time in the manuscript.
- Line 105, please use relevant references for the sentence “Strict rules on ……..” and the following sentence.
- Line 128, please use latitude and longitude information of study area. Also, show in the map in Figure 1.
- Line 131, please use relevant references for the sentence “Life expectancy …”.
- Line 138, please use relevant references for the sentence “the highly productive …”.
- Line 139, please use relevant references for the sentence “Additionally, urbanization …”.
- Line 188, please use only VIs, since you used the abbreviation earlier in the manuscript.
- Line 193, please use the abbreviation NDVI and LSWI in bracket.
- Line 193, please use relevant references after “…. in previous studies”.
- Line 256, please use relevant references after “…entire country.”
- Line 269, please use relevant references after “…. earlier studies.”
- Line 292, please define BBS before using as abbreviation.
Author Response
Dear Reviewer,
Please see our point-by-point response attached in the pdf. The authors greatly appreciate the efforts of the reviewer to improve our manuscript. We believe your critiques and suggestions, along with our responses, have enhanced this research.
Thank you.
Author Response File: Author Response.pdf
Reviewer 3 Report
The authors provided a well-structured manuscript presenting their study in Bangladesh for identifying dry season rice patterns using the Landsat archive (2014-2018).
However, I have two main issues with the presented methods and validation of the results. The authors have to better clarify:
Why they have used the TOA Landsat data instead of SR? The authors provided some explanation in Lines 216-218, but in my opinion, it would be interesting also to use the SR Landsat archive data for the same period and area and compare the results. This also because the EVI index that the authors used for this study could be affected by atmospheric effects not properly corrected using the TOA reflectances.
I understand that the validation for such a wide area is very difficult, but the obtained model should be validated and improved also by comparing results with different in situ information so to understand the potential of this rice-mapping technique. The authors should improve section 3.4. to confirm their R2 values. As they mention in section 3.4., the BBS data are estimates gathered based on local extension offices, regional agricultural data collection centers, and sampling surveys of farmers in each region. The data are aggregated areal estimates at the district level and do not provide explicit spatial details or location-specific management details. It would be nice to have some more accurate ground validation performed on some spot/test areas to confirm and validate their results.
Minor comments:
Line 56: add some other references, e.g. Busetto, L., Casteleyn, S., Granell, C., Pepe, M., Barbieri, M., Campos-Taberner, M., ... & García-Haro, F. J. (2017). Downstream services for rice crop monitoring in Europe: From regional to local scale. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(12), 5423-5441.
Line 73: the verb is missing.
Line 138: add the reference number
Figure 1: the caption is incomplete and Figure 1c is not pertinent here and of scarce quality.
Figure 2: move the description of A and B in the caption and not on the figure, also without mentioning the positions (left and above).
Lines 238-239: what represents alfa and which values did you assigned in your work?
Lines 261-264: use the same verb form as in point i).
Lines 275-276: rewrite as..."we the compare phenology information of rice with the wheat growing plots".
Figure 3: rewrite the figure caption including a description of A, B, C and D.
Author Response
Dear Reviewer,
Please see our point-by-point response attached in the pdf. The authors greatly appreciate the efforts of the reviewer to improve our manuscript. We believe your critiques and suggestions, along with our responses, have enhanced this research.
Thank you.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
The authors have substantially improved the manuscript following also my suggestions and comments. In my opinion, the paper could be published in the present form taking into account that some improvements, especially in the validation phase, should be made.
