This paper investigates the cause of long-distance Automatic Identification System (AIS) links (>1000 km) observed in the western Atlantic region during 13-14 July 2021. It is hypothesized that the long-distance links are enabled by skywave propagation due to sporadic-E layers. Digisonde data and COSMIC-2 radio occultation measurements support this hypothesis. It is suggested that AIS data could be a new tool to study intense sporadic layers.

The paper is written well and contains an interesting case study. My overall opinion is that there is a merit of publishing this paper, as it is the first report on the potential use of AIS data for global sporadic E studies.

Nonetheless, I have a comment, which the authors might want to take into account for improvement of the paper. It feels that the discussion of the sporadic-E maps in Figure 5 is incomplete. The maps are presented only for two selected intervals on 13 and 14 July, and it is unclear how they look, for example, on 15 July or nighttime on 13 and 14 July when long-distance links were not detected. It would be informative if the authors could demonstrate the absence of intense sporadic E. In case there is no sufficient data for constructing maps for other times, the authors could simply state so.

The paper is about the detection of long distance AIS signals in the western Atlantic ocean, supposedly propagating via sporadic-E layers.

The paper is nicely written, I definitely enjoyed reading it and support its publiction.

My main complains are

1) the statements of tropospheric ducting being limited to roughly 1000km, which is not correct. Indeed, without any distinct inversion/ducting the tropospheric link distances should be in the order of a few 100km as correctly noted in Line 25.

There are many trustworthy reports of bridging several thousands of km in the amateur radio community on 144MHz and above, e.g.:

https://mmmonvhf.de/odx.php

Like the currently reported largest distance on 144MHz tropospehric ducting:  Canary Islands <-> Puerto Rico, nearly 5000km

There are tropospheric ducting forecast maps available online, e.g. 

https://dxinfocentre.com/tropo_car.html

Perhaps the statements could be relaxed regarding above 1000km can't be tropospheric ducting.

Hence, even though the Digisonde and COSMIC data is very convincing, maybe it would be possible to rule out strong tropospheric ducting by tropospheric temperature inversion maps for the 2 days in question?

2) I also have little doubts of interpreting the 1000km distances as sporadic-E propagation, this is more typical for lower frequencies, say 50MHz. Otherwise the foEs must have been indeed extremely high at that time. But these are just a few ten examples at the end in your histogram and the general distribution indeed indicates 1000km-ish for the given days.

3) In Line 54 you indicated a maxium distance of 5463km, I'd find it interesting to see this path in the given maps. I'm confident it's been a double-hop Es.

Technical issues:

- inconsistence of   value and units , specifically for distances - L 10, 22, 42, 46, 47, 51,....

- caption Fig.1     13 -15 July    -> 13 - 15 July

- E.g. in Line 72 explain/note the abbreviation RO, it's not clearly introduced earlier

- Lines 72/73, I suggest to add the coordinates for the Digisondes or quote they are marked in Fig. 5

- Fig. 4 - y-axis label of the middle panel: WALLOPSIS - I'm sure it's Wallops Island, but perhaps just using WALLOPS there? Furthermore should the label be f0Es not FoEs as used in the text, right ? Same for the legend.

- Line 94: Does that mean, the shown intensities are scaled, or are adjusted to the measured foEs by the Digisondes? Perhaps rephrasing would be good...?

- Line 104/105: Could you give a reference of the secant law to calculate the f0Es or necessary electron density?

- Lines 108-114: Is this just a quote, proposal for further improvements or did you already apply such methods in this study? It's not strictly clear to me.

- References: looks like some DOIs are missing, I propose to add them.

Again, I support the publication of this manuscript after adressing the mentioned points.