Please see attached comments.

This paper attempts to cover the topic of the impact of convectively coupled tropical waves during the CADDIWA field campaign on atmospheric circulation, water vapour, dust, thermodynamical vertical structure, and tropical storms in the vicinity of Cabo Verde and the west African coast. The focus is on the period of the field campaign in September 2021, but use is also made of some longer periods of data to make more general conclusions of the wave impacts (2018-2022 in Figure 1, July-September 2021 in Figure 3, 2003-2021 in Figure 5). Of the waves considered, they find that Equatorial Rossby (ER) waves are the most significant driver of water vapour variability near the equator and north of 20˚N while the combined mixed Rossby-gravity (MRG) and Tropical Depression (TD) type of waves have most impact near the location of the African Easterly Jet (AEJ) around 13˚N. They also present evidence for variations in dust and an association with four named tropical storms (Larry, Pierre-Henri, Peter, and Rose). Data issues were also discussed.

Unfortunately, I found the paper to be quite a challenge to read. I think it could be improved if it was shortened with greater focus on the more important results and perhaps spreading the current content across more than one paper. I also found some of the analysis and results confusing, as detailed below. I hope the authors can consider these points to generate a new and improved manuscript for journal submission.

Specific comments:

Abstract, line 1 and elsewhere. In most previous on tropical waves the acronym TD has referred to Tropical Depression, not Tropical Disturbance.

Abstract, line 3. What does it mean for the MRG-TD tracks to be "mingled in the literature"?

Abstract, lines 11-14. The title and the previous text of the abstract suggest that the paper is about dust and atmospheric thermodynamics only, so this inclusion of a discussion of tropical cyclogenesis seems off-topic.

Introduction, line 87. You say that you follow the method of Wheeler and Kiladis (1999) but the "protocol" of Janiga et al. (2018). Please include a sentence saying how the method of Janiga et al. differs from Wheeler and Kiladis (1999).

Section 2, line 129. What do you mean by the "Real" shallow water model?

Line 139. The important point to add here is that the wavenumber-frequency filtering will output wave-like structures from input that is purely red-noise, so it is not always straightforward to attribute the output to a theoretical wave.

Figure 1a. In the introduction you said that you used the protocol of Janiga et al. (2018), but they used different regions of wavenumber and frequency for filtering as displayed in this figure. What did you use?

Figures 1b and 1c. The period 2018-2022 is clearly not long enough to get a clear picture of the waves in the spectrum. I suggest you use a longer period. Also, does your spectrum use data from all longitudes, or is it focussed on the Africa-Atlantic region?

Line 158. Say: "Figure 1 shows that the wavenumber-frequency domains of MRG and TDs do overlap significantly".

Line 191. Here you say that you base the composite horizontal structure on the TCWV data averaged for 0-20˚N but at line 170 you said 0-15˚N. Which is correct?

Figure 3. I am confused by the composites presented in Figure 3. Why aren't the TCWV structures more symmetric or anti-symmetric about the equator? If you do the wave filtering using the symmetry constraints of Wheeler and Kiladis (1999), as implied by Figure1bc, then I expect that the TCWV composites should be more symmetric (or anti-symmetric) than what they are.

Figure 3cd. I see nothing in this composite structure that resembles a theoretical Kelvin wave. I think you are instead looking at an artefact of the red noise background. My conclusion that it is an artefact is supported by Figure 5 that shows the Kelvin wave signal is maximized at about 30˚N, which is unlike the theoretical Kelvin wave structure which should be within about 15 degrees of the equator for the equivalent depths that you are considering.

Figure 3 caption. You say that the number of events for the composites is included in the titles, but I don't see that.

Figure 5. As I said above, I do not trust the results presented for the Kelvin wave. I think it is an artefact of the method.

Figure 5. The caption says it is the "Relative importance of tropical wave signals". Rather, it is the Squared correlation coefficient between the filtered wave signals and the associated variable. It should be noted that this is likely an overestimate of the relative importance given that there is likely a lot of noise that also contributes to the variability in the wave filtered fields.

Figure 6. Once again, the location of the peak of the Kelvin wave "relative importance" does not make physical sense with regards to the Kelvin wave structure which should be centred near the equator.

Figure 7. You have labels for tropical storms Rose, Peter, and Pierre-Henri, but you do not show their actual genesis time/location or their track, just a vague timing.

Line 344. You mention TS Larry here, but it is not shown on Figure 7.

Conclusions versus title. Much of the conclusions focussed on the tropical storms and TC genesis which was not reflected in the title.