A Novel Method for Cloud and Cloud Shadow Detection Based on the Maximum and Minimum Values of Sentinel-2 Time Series Images

Round 1

Reviewer 1 Report

The paper proposes a novel method for cloud and cloud shadow detection based on Sentinel-2 time series images, which utilizes the temporal maximum and minimum mask method (TSMM) to estimate the temporal features and incorporates the spatial convolution operation to fully leverage the complementary advantages of temporal, spatial, and spectral information. The method is tested on two datasets, the national S2ccs dataset (manual) and the global CloudSEN12 dataset, and achieves higher accuracy and F1 scores compared to five advanced methods or products. However, there was a small number of questions

1.The abbreviation and spelling of TSMM do not appear in the abstract.

2.The method mentioned in section 3.3 about salt and pepper noise, which may limit its applicability.

3. The method's performance under varying atmospheric conditions or in different geographic regions.

4. How the method compares with the latest deep learning approaches, especially in terms of accuracy and generalizability.

5. In Section 2.2, what criteria were used to select the 7 locations for conducting experiments on the dataset? Are these datasets sufficiently representative to ensure the universality and reproducibility of the experimental results?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper introduces a novel method, temporal maximum and minimum mask method (TSMM), for detecting clouds and cloud shadows in Sentinel-2 time-series images, utilizing maximum and minimum values in the blue and near-infrared bands, respectively, to improve detection accuracy. This is a topic with practical application value, and the experiment design is reasonable. However, there are some problems in the article that need to be slightly modified.

My major comments are:

(1) The abstract section is overly verbose and wordy. Please consider refining the text of the abstract to make it more concise.

(2) The content of Table 1 consists of basic knowledge in the field of remote sensing, which might appear too elementary for submission to a specialized remote sensing journal.

(3) Figures 7 and 8 are too large; please select representative results for display or place the images in supplementary files. This will enhance the reader's experience.

The language of the article requires minor improvements.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf