Review: Brief Communication: Accurate and autonomous snow water equivalent measurements using a cosmic ray sensor on a Himalayan glacier by Pokhrel et al. 

The study presents an interesting combination and comparison between SWE measurements by a cosmic ray sensor (CRS), precipitation measurements and model simulations by COSIPY in the lower accumulation area of a Himalayan glacier in Nepal. With their measurement setup, the authors draw important conclusions on processes in the accumulation area of a glacier. The overall results and conclusions are very interesting and worth a publication. However, revisions are needed to provide more clarity on the measurements and the conclusions and to improve the readability of this manuscript. 

The following open questions/ points should be addressed: 

1. The title is somewhat misleading. The title gives the impression that the study presents a thorough evaluation of the cosmic ray sensor in these high-altitude areas. Yet, there are only three manual measurements presented as a reference at the same site (of which one has no corresponding CRS observations (see Fig.2). For the title to be less misleading, the word “Application” could be integrated (or something in that sense). 

2. It is not outlined how the authors dealt with the main issue of deploying a CRS in an accumulation area of a glacier: Is the device dug out every season, or will it slowly be buried in the glacier until it does not work anymore? What are the absolute neutron count numbers and their evolution over time (maybe add such a plot in the supplement)? What are the uncertainties of the CRS and how strongly do the counts fluctuate over their time intervals?

3. A Brief Communication should be kept short, but it would still be helpful to dedicate two or three sentences to the COSIPY model given the importance of this model for this study. 

4. The authors compare delta SWE derived by CRS measurements to the simulation of mass fluxes by the COSIPY model, but this evaluation is only done qualitatively even though approximately two years of measurements are available. In addition, uncertainties of the CRS measurements (or the COSIPY model) are neither taken into consideration nor discussed.

Detailed comments: 

L22: Gugerli et al. (2019) present a performance assessment of the CRS and information that can be gained from such measurements. I recommend to use a more suitable references for such a general statement. 

L26-27: Please revise this sentence. 

L40: Please note that neither Howat et al. (2018) nor Gugerli et al. (2019) measured 4000 mm w.e. above the CRS. Gugerli et al. (2019), e.g., measured approximately 2000 mm w.e. of SWE (and snow depths up to 5m). If this estimate of 4000 mm SWE is taken from Fig. 3 in Gugerli et al. (2019), it corresponds to a theoretical estimate of what the sensor should be able to measure, but this has not yet been demonstrated (at least not to my knowledge). 

L84: How did you define the depth at which to bury it? Did you excavate a pit until the first change of snow density (which is the depth of 20cm you mention)? How was the snow/ firn below that? 

L86: Were the counts averaged or summed over the hour? In Howat et al. (2018) and Gugerli et al. (2019), for example, the counts are summed over an hour (cph). It would be interesting to keep the way the counts are presented in the same way to more easily allow for a comparison across different sites.

L115: It is not very clear how you obtained the reference count. Above you write that the scanning time was 20 seconds, and these counts are averaged over an hour resulting in counts per hour. For the reference count, however, you average 1-minute counts. Wouldn’t it be more consistent to use the same measurement strategy in both cases, also to use the same reference period for all the correction factors? 

L127: “through good agreement with field measurements” – According to Fig. 2, there are only two field measurements available which can be compared to the SnowFox measurements, and the first one appears to be obtained during the deployment of the SnowFox and hence with a disturbed snowpack.  

L130: What does the uncertainty of the manually measured SWE respresent - a standard deviation of several measurements, or an error propagation? 

L135: How long was the second gap?