|The authors modified and greatly improved their manuscript. I suggest publication of the paper after only two minor corrections which are not altering the general quality of the results. |
1. In their response, the authors write that in the analysis of Libois et al. (2014), sastrugis are associated to high peaks in snow height measurements lasting only a few hours, which should disappear if a running mean of 24 hours is applied on the hourly snow height measurements. => if the authors look more accurately to Figure 2 from Libois et al. (2014), with a focus on surface height variations in august/September 2012, they will observe that the sonic gauge proposes 1) a significant snow height increase, 2) constant values during a few day 3) a significant decrease of the surface (erosion). These events are probably due to the movement of sastrugis, which likely stopped moving below the sonic gauge during a few day and then was eroded again. The authors acknowledge that 3 reptation events may impact their results but conclude that: “Apart from this, temporary peaks (i.e. with a strong decay within 48 hours) in the snow height records are excluded from the analysis”. I suggest that the authors smooth their comment and write that a 24 hour running mean should remove the “majority” of peaks associated to the motion of sastrugis. Nevertheless, few events may also be related to artefacts caused by the motion of sastrugis.
2. the authors write that “Since the magnitude of the two sublimation terms is small over the PE station, the uncertainty on the ERds component is mainly determined by the uncertainty of the snowfall estimate”. However, the cumulative Sublimation and ERds values are in the same range of magnitude. As a consequence, significant uncertainty in Sublimation can impact significantly the ERds values. Many papers discuss the uncertainty of surface sublimation (e.g. bliss et al., 2011). Getting the uncertainty of the SUds term is more complex but may be discussed using results from RACMO2.3 model => I suggest that the authors indicate an uncertainty range for both sublimation terms, or give values (or statistics) of each term for several typical events demonstrating that ERds is determined by snowfall estimate.
Bliss, A. K., Cuffey, K. M., and Kavanaugh, J. L.: Sublimation and surface energy budget of Taylor Glacier, Antarctica, Journal of Glaciology, 57, 684–696, https://doi.org/10.3189/002214311797409767, 2011.