Publisher’s note: a supplement was added to this comment on 24 April 2023.

Pine Island Glacier is a major contributor to the loss of ice mass. Work to date has focused on understanding basal melting. The authors use 41 years of reanalysis data to investigate the role of Foehn winds in the surface mass balance at Pine Island Glacier. They characterise Foehn episodes over the satellite record and show that sublimation plays an important role in surface mass balance. Sublimation processes are not captured existing remote sensing techniques and this article highlight an important contribution to ice mass balance that is currently not monitored. 

The article is well-written and well-structured. It is an interesting article that provides a new aspect to our understanding of how Pine Island Glacier is changing in defining the role of Foehn winds in surface melting. My comments are largely related to improving the readability of the figures and the length of the article. 

My one main concern is around confidence in  the ERA5 dataset. The authors note that ERA5 does not have a sufficient spatial resolution to fully resolve smaller-scale flows. However, they also note that previous work has shown that moderate to strong Foehn events can be identified over the Antarctic Peninsula. Does this location of the AWS allow an assessment of the proportion of Foehn events at PIG that are resolved in ERA5? If this could be shown, it would lend more weight to the conclusions. 

It is a long article, and the authors might consider some of the following suggestions to reduce the length of the article:

• Introduction - you later show that SAM does not play a large fole in the weather conditions at PIG, so perhaps you don’t need to include the details on SAM in the introdution (lines 116-121). 
• Less emphasis in the introduction on Foehn processes around Antarctica and instead focus on an introduction to the important details in and around Pine Island Glacier. Similarly in section 3, the Foehn events at Pine Island Glacier are extensively compared to Foehn events elsewhere. Some of the details from other locations can be more concisely presented and the focus kept on what is happening at Pine Island Glacier.
• Lines 334-345 there is a long discussion about trends. With the amount of variability in the signals, the timeseries is too short to identify a trend. Consider just pointing out the small trends and low statistical significance

Minor comments

Figure 1a: it is very difficult to differentiate between the thick and thin purple lines - perhaps consider using two colours. I find the purple quite hard to read on the dark green so maybe a different colour scheme would also help here. Perhaps you could show the location of the Evans Knoll AWS on Figure 1a. It would also be help to orient the reader if you could show the location of the outline for Figures 3 and 4 on this figure - I don’t think they are the same as the red box shown?

Figure 2a and b: I found it tricky to related the yellow boxes to the red axis and it took me a while to work out how to read these graphs. Consider using the same colors for the bars and/or adding a legend as you do in Figure 2c.