This manuscript presents ice shelf cavity observations in Pine Island Bay. The value of these observations is indisputable, and they deserve to be published as proposed in this manuscript. I do not have any major criticism. If any, I wish the presentation should be improved to better highlight the novelty of the results, and to provide a slightly improved description of methodologies.

Novelty: Reading the abstract only, I find it hard to appreciate what makes this study novel and unique. Could this be restated and improved?

The map in Fig. 1 was so little that I found it hard to read. The choice of colors could also be improved.

DTS thermal profiling: I could not find a definition of the acronym DTS before l. 144. It would be nice to find information about the accuracy of these measurements. Is there a way to assess the potential for temporal drifts?

Cross-wavelet analysis: I found the amount of information available on this method insufficient. What is the unit of the quantity derived from the cross-wavelet transform? How to interpret the result? More generally, would it not be useful to see a wavelet transform of the temperature signal alone? Also, the correlation between temperature and salinity implies some level of density compensation. Would it be possible to analyse the density variations directly? It is partially done in Fig.  6 but it could be better highlighted.

l. 390: I do not understand the statement that warming leads to thermal expansion of the water column. The direct effect of thermal expansion on the position of an isopycnal would be at a centimetric scale at best, especially in such a cold region. This is not something I expect can be directly observed. Can you clarify?

Fig. 8: I find it very hard to understand what the x-axis corresponds to and how to read this figure. More details would help.

The richness of the datasets and material presented in the Figures here is appreciated, however one easily loses track of the correlations in time between T,S,sea ice and current data, and of the bigger picture conclusions, especially if one is not familiar with the region.

I suggest the following: 1) the figures and their content could be introduced more properly (often times a relatively complex result is stated and then just "Fig.x" in brackets, and the reader has to go to the figure first, read the caption and then try to make sense of it) 2) introduce more subsections with goal of the subsection and make a "big-picture summary" at the end of each subsection. 3) mark relevant times in the Fig. so one can keep better track of what is happening when in different variables 4) a summary sketch at the end summarizing bigger picture events 5) a better and bigger map at the beginning to show what is relevant

In the introduction, can you specify more clearly what is different from previous work and which additional information you will analyze here.

mark in Fig.2 relevant time periods that are emphasized in the text to be able to keep track

solid line is not the mean

line 228: at depth, the warming trend starts more like in August 2020, before that there is a cooling trend?

line 240 : I don't really see the warming trend at depth from April on, more from August 2020 on?

line 262: from Fig. 3 it is not so clear that co-variation happens at all periods. Even though the arrows show co-variation

the significance clearly drops sharply with small periods.

just visually from Fig. 2, co-variance at long time scales is not evident for the Channel Camp 316m, there is a clear warming but no freshening trend

Fig.3, Fig.6 add colorbar

line 277: to be able to keep better track of the distinct events, somehow marking them in Fig.3 would be good,

and relate them back to the events/trends evident from Fig.2

line 289: it seems that something new is discussed now, so a new subsection would help, together with a brief intro on what Fig.4 shows to help the reader digest the information

line 308: was the "Gade line" introduced before?

line 353: something important and new is starting here, so a subsection would be good. I like the introductory sentence as motiv

ation, something like this should appear at the beginning of each new subsection.

line 367: something new starts again

Fig.7 would it make sense to somehow plot differences in time or wrt the mean to emphasize trends, the lines are all very close together, mark relevant depth intervals mentioned in the text?

is Fig.10 i,j,k discussed?

Fig.11: I appreaciate it, since it shows a big-picture schematic summary, but please properly take the reader through Fig.11

it would be great if at the end there could be a summary sketch, summarizing the major findings and conclusions that were evident in all the figures thereby bringing them together