See attached file

This paper aims to produce a dataset of Chinese rivers spanning the period from 2016 to 2023 at an annual scale with a resolution of 10 m. However, the dataset lacks originality and has gaps in sufficient quality, and is limited in its potential for broader application, which I detail below:

Originality:

1) The classification of water body is more easily achievable compared to other land cover types in the field of remote sensing. It exhibits a significant spectral difference from other land cover types and has a relatively simple texture. Furthermore, it would be easy to screen rivers by simply using the length-to-width ratio of water bodies. However, the authors utilized publicly available 10 m land cover data and did not used an innovative scheme to extract rivers. They also fail to consider the network of rivers and the topographical features that influence the formation of rivers. The originality of the technical solution is limited.

2) As shown below, the authors did not acknowledge many relevant river datasets in the text. This makes me seriously concerned about the proper place for this paper.

Lin, P., Pan, M., Wood, E. F., Yamazaki, D., & Allen, G. H. (2021). A new vector-based global river network dataset accounting for variable drainage density. Scientific data, 8(1), 28.

Nyberg, B., Sayre, R., & Luijendijk, E. (2024). Increasing seasonal variation in the extent of rivers and lakes from 1984 to 2022. Hydrology and Earth System Sciences, 28(7), 1653-1663.

Yan, D., Wang, K., Qin, T., Weng, B., Wang, H., Bi, W., et al. (2019). A data set of global river networks and corresponding water resources zones divisions. Scientific data, 6(1), 219.

Besides, the work of Allen & Pavelsky (2018) has been cited, but the differences from the data of this article have not been explained.

Allen, G. H., & Pavelsky, T. M. (2018). Global extent of rivers and streams. Science, 361(6402), 585-588.

Scientific quality:

1) The scheme of data validation has considerable uncertainty. The river is typically characterized by the property of network morphology. However, the generated river data display a substantial number of river discontinuities. There is a conspicuous phenomenon of rivers in adjacent years either "disappearing" or "breaking off" noticeably. Although the accuracy is about 95% by visual interpretation, this is based on a pixel-by-pixel basis and does not take into account the connectivity of rivers. Additionally, the visual interpretation is also a highly subjective process. If only the center of the river was selected, the accuracy of the river would be overestimated.

2) The key data utilized in this paper (i.e., European Space Agency and Dynamic World) are 10 m land-cover classification products. They were not primarily designed for water classification. These products tend to underestimate the area of water body, and consequently, the extent of rivers.

3) Rivers possess highly pronounced seasonal characteristics. During the summer flood season, rivers become wider, while in winter, they may even disappear. The specific meaning and significance of annual-scale river data remains unclear.

4) "For areas with missing DW data, the EGLC and Sentinel-2 images were chosen as supplementary datasets, which were utilized to create annual water maps." This strategy is subjective. As depicted in Figure 2, the EGLC data only encompasses the period from 2017 to 2023. In contrast, for the remaining years of 2015 - 2016, classification is carried out using the land cover data that was self-produced. Why use the DW dataset as the primary data of river extraction? How to ensure consistency across datasets? The experimental scheme also has a certain degree of subjectivity.

Application

1) As illustrated in Fig. 8, the river data produced in this paper is significantly different from that of other products. In practical applications, it is relatively difficult for users to make a trade-off regarding which one to use.

2) This is not a global product. It has relatively limited application potential compared with other global river products.