Multi-Mission Satellite Detection and Tracking of October 2019 Sabiti Oil Spill in the Red Sea

Round 1

Reviewer 1 Report

The manuscript 'Multi-Mission Satellite Detection and Tracking of October 2019 Sabiti Oil Spill in the Red Sea' presents a multi-mission satellite remote sensing approach to detect and track oil spills. SAR images (from Sentinel-1) and optical images (from Sentinel-2, MODIS, and Landsat-8) are combined and used together to detect oil spills. The inferred oil spill movements are validated to be consistent with the ocean currents. There are some comments and suggestions as follows:

1. All formulas should not have two numbers after them.

2. The URL for downloading these remote sensing images should be given.

3. The manuscript only shows the oil spill detection results, but how to detect oil spills from remote sensing images is not described.  With the development of deep learning, oil spill detection based on CNNs has been studied widely such as 'Oil spill detection with multiscale conditional adversarial networks with small-data training' and 'Oil spill identification from satellite images using deep neural networks'. These CNN-based detection methods play an important role in the oil spill detection area. In the future work, it is suggested to utilize these methods to detect oil spills from remote sensing images.

4. The manuscript only validates the inferred oil spill movements are consistent with the ocean currents as revealed by a high resolution regional ocean reanalysis. In the future work, it is suggested to utilize a numerical prediction model to predict the oil spill movements such as 'An adversarial learning approach to forecasted wind field correction with an application to oil spill drift prediction'. The oil spill movements are driven by both wind fields and current fields, and the single current fields can not describe the drift comprehensively.

Author Response

Please see attachment

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript "Multi-Mission Satellite Detection and Tracking of October 2019 Sabiti Oil Spill in the Red Sea" is a good case study showing the capabilities of each available satellite-based sensor to detect and track oil pollution on the sea surface. The described material can be useful as an illustration in teaching students about satellite monitoring techniques and for popular scientific purposes.

For scientists dealing professionally with the problem, this article is of interest, but the methods described in it are well known.

I would like the authors to try to estimate the thickness of the film at different times; provide an explanation why the oil film remained on the sea surface for so long time; and write whether attempts have been made to stop the spreading of the oil and to collect it from the sea surface

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

I think the paper is fairly well written and presents a comprehensive approach to multi-modal oil spill tracking, but it could do with a bit more detail at points and some improvements in presentation.

1) This one is fairly minor, but you may want to include in the beginning of Section 2 where Sentinel-1 is described, that Sentinel-1B is no longer operational. It went out in early 2022 and a few days ago the decision to stop trying recovery was made, so it's gone for good. This means revisit times are back at 12 days until Sentinel-1C is launched, which is going to be a few years from now. Makes the point of SAR having bad revisit times even stronger. 

2) In Figure 1 showing your area of interest; I'd suggest you include a scalebar on the map, and possibly a zoomed in version of the region of interest showing the location and extent of the oil leak (and possibly the exact coordinated of the green start point you have marked out). 

3) The Tables throughout can probably do with some better typesetting, and avoiding to split them over multiple pages.

4) For Sentinel-1 IW GRD products, the pixel spacing is 10m; resolution is around 20m and in SAR is more used to clarify the distance between two separately resolvable targets. It might be good to clarify you're referring to pixel spacing. 

5) The oil spill detection method for SAR could do with a bit more detail; are you simply thresholding anything between -18 to -35 dB? How are those boundaries determined and how do they change depending on sea state? I'd suggest you go into a bit more detail on how oil spill detection on SAR is performed, especially as you often use the outlines derived with these in later evaluation/comparison. The methods you have used for other modalities (like OSI for Sentinel-2) are pretty straightforward and well established, but for SAR it is not fully clear how you detected oil spills.

6) Figure 2: This is a good opportunity to include more details about the SAR SNAP processing. I'd suggest you incorporate some of the important parameters in this graph. Multilooking for example - how many looks did you take in range/azimuth? This is important, as it will determining the pixel spacing of your end image, on which you do your classification.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 4 Report

The methodology of satellite image processing is adequate and well explained. The authors used available sensors from various satellites (SAR, visible, IR) to detect the Sabiti Oil spill and its spreading in the Red Sea. They also used model reanalysis of currents.

It seems that this oil spill event is well documented regarding the satellite data, however missing are the wind data. Particular wind conditions are crucial for detection of oil at the water surface as well as for its spreading. They should include some wind observations either from coastal stations or retrieved from satellites. It must be possible to find some wind data.

In the figures 6, 12 and 13 arrows of currents are hardly visible, better to show less but longer arrows, or at least mention the range of speeds of currents at these images.

It is not explained are there in that period only thermohaline currents or some currents were wind driven?

I did not see here that the authors mention the quantity of the oil released in this accident? There must have been some environmental effects due to this long duration oil spill event, but it has not been mentioned neither in citations.

Author Response

Please see the attachment

Author Response File: Author Response.docx