MFF-Net: Flood Detection from SAR Images Using Multi-Frequency and Fuzzy Uncertainty Fusion

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents MFF-Net, a multi-frequency fuzzy uncertainty fusion network designed for flood detection in SAR imagery.  While the topic is relevant and timely, several issues should be addressed to improve clarity, completeness, and scientific rigor:

1. Ensure that each abbreviation is defined upon first appearance and used consistently thereafter to improve readability for multidisciplinary audiences.
2. Prior works combining SAR–optical, multi-temporal, or frequency-diverse SAR data should be discussed literaturly.
3. An ablation experiment comparing “with vs. without” the fuzzy uncertainty module is necessary to demonstrate its quantitative contribution to noise suppression and fine-grained flood area detection.
4. In the later sections (starting from Section 3.6), the study compares the proposed method only against two baseline models. Given the wide range of flood detection algorithms in recent literature, this limitation should be justified.
5. Figures 12 and 13 indicate that MFF-Net exhibits the highest omission ratio, suggesting more missed flood pixels compared with other methods. However, Tables 7 and 8 show that the proposed method consistently achieves the highest IoU/F1/OA/Recall values. This apparent contradiction should be explained. 
6. A key practical challenge in operational flood mapping is distinguishing flooded areas from permanent natural water surfaces. The abstract claims that the proposed method detects fine-grained flood areas effectively, but it does not explicitly address whether MFF-Net can separate floodwater from background water bodies. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

• In Tables 1 to 5, some numbers are in bold and others are underlined. I understand that the former are the best results and the latter are the second-best results. Please clarify this in the text.
• In Tables 6 to 12, there are numbers in bold and no underlined numbers. To maintain the same format for displaying results, I suggest underlining the second-best results.
• When the results are shown in tables (Tables 1 to 12) for the different chosen metrics, the MFF-Net method consistently performs better than the methods with which it is compared. Clarify this capability in the text of the manuscript, as it makes MFF-Net more robust than the other methods.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Please find the comments in the attachment. 

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf