It is intresting about the d-CNRS for deep soil water measurement through groundwater wells, but I only suggest the manuscript should be shorted and the story wil be easy for reading.

The study is very interesting and the new setup of downhole cosmic-ray neutron sensing for the estimation of soil moisture within vadose zone looks promissing. As it is mentioned by the authors further research is definitely needed to test the proposed approach in different sites. 

It would be good to see the estimated soil moistures from above ground CRNS in Figure 10.  How different the values measured by above ground CRNS and downhole CRNS at 100 m depth?

Is the precipitation in liquid form during the winter? Does snow affect the downhole neutron intensities? 

The authors introduced a novel way of using the cosmic-ray neutrons sensing technique.

They modeled and measured thermal neutron radiation at below surface dephts, to estimate soil moisture at depths greater than the measurement depth of the more traditional CRNS soil moisture sensor (~40 cm below surface)

To make use of existing monitoring infrastructure, they put CRNS sensors in groundwater level observation wells, at different dephts, up to 10 metres below the surface.

When measurement time was integrated to 25 to 49 hours, sufficiently precise measurements could be obtained up to ~2 metres below the soil surface.

The somewhat eliptical measurement support volume came out larger than that of traditional active below-ground neutron soil moisture sensors, with a radius of ~50 cm at wet conditions.

To avoid calibration against in-situ soil moisture sensors or sampling, they used scaling against above-ground thermal CRNS measurements over a water body.

The authors acknowledged relevant uncertainties on the measurements and their research outcomes and provided a look forward.

Especially regarding the calibration method without soil moisture observations, more research needs to be done, according to the authors.

Moreover, the specific results were deemed valid for the specific sites only, in principle.

To be certain the outcomes can be used at different sites, more research is needed.

The authors wrote their manuscript with a clear structure.

The research is novel and relevant to the HESS audience and is written in a way this audience should be able to understand, even with limited knowledge of particle physics.

I recommend a few minor changes only, indicated hereinafter.

- In the introduction (Chapter 1), the publication by Kodama et al. (1985, https://journals.lww.com/soilsci/Abstract/1985/10000/APPLICATION_OF_ATMOSPHERIC_NEUTRONS_TO_SOIL.1.aspx) should be mentioned.

This publication is about a passive below-ground neutron sensor and therefore it is needed to briefly mention how the current manuscripts relates to this work from the 1980's.

- Chapter 2 Material and Methods, Section 2.4, page 9, lines 196-204: The HESS-reader could be helped by explaining a bit more how the roles protons, muons, and neutrons play, differ and how much different processes dominate in relative terms (moderation and in-soil neutron production)

Please, if explained in the manuscript text, keep this very brief.

- Chapter 3  Results, Section 3.1.3, page 7, line 362: Why from wilting point to field capacity (these are rather arbitrary soil moisture contents that, formally, relate to the plant grown) and why not to saturation?

- In chapter 4 Discussion: The possible uncertainties introduced by vertically highly heterogeneous soils should be discussed.

These heterogeneities can affect the outcomes through both vertically variant soil bulk densities and, during certain periods, vertically heterogeneous soil moisture contents.

To what extent does not knowing the exact vertical variation matter to the outcomes and uncertainty?

To what extent would vertically heterogeneous soils affect the applicability of the methods at different sites, given more in-situ sampled data might be needed?

- Chapter 4 Discussion, page 23, line 466 "...may lead to different results": such as?

- Chapter 5 Conclusions, page 27, lines 585-590: Please, discuss here and possibly in Chapter 4 Discussion, Section 4.2, that using soil hydraulic models could introduce further uncertainties due to assumptions made during the modelling process. In addition, applying a soil hydraulic model to obtain the results, make these less observations and more model results.