1) Originality (novelty): 2

In this contribution, the authors present a laboratory test using ground-penetrating radar to monitor the active layer dynamics. From the reviewer’s understanding, the novelty of this contribution is that the authors used thermal and volumetric water content sensors as reference sensors in the experiment to understand the freezing/thawing process better.

2) Scientific quality (rigour): 3

The experiment is well constructed under laboratory conditions. Section 2.1 is well written, except it is unclear whether three or four volumetric water content sensors are used.

The reviewer doubts if bedding the bowtie antenna by 30° can help to focus the energy.

Equation (3) does not match its description (Line 91). The reviewer cannot understand why neither \epsilon_w nor \theta_i is given in equation (3). Further, in Section 3.2, the reviewer cannot understand how the dielectric permittivity distribution is derived from equation (3).

3) Significance (impact): 3

Since the main result is poorly presented, the reviewer cannot judge the significance of the contribution.

4) Presentation quality: 4

1. Poor writing style with many typos. The notations are not consistent, for example, S11 and zero-isotherm.
2. Mistakes, e.g., line 103: “the depth of the zero-isotherm reaching the ground surface after 125h from the start…” does not match figure 2c).

Line 145: “efficiency …is…less than a centimeter”. What does this mean?

1. The quality of the figures is poor. Every figure has another font size. The axis label of Figure 3(b) is even half covered. The reviewer cannot understand Figure 3(b). If we look at the t=0 curve, does it mean at the beginning of the experiment, the height of the sand is only about 0.1 m? Figure 3(d) should be an important result of the contribution. However, it is very poorly described. What are the black dots in the plot? Why are they not been used for calculating the linear regression?

Geophysical methods, and in particular Ground Penetrating Radar (GPR), have been used extensively to study permafrost and frozen subsurface environments, mainly through ad-hoc surveys, i.e. there has been little previously reported on in-situ GPR sensors for time-based monitoring over extended periods. The authors look to address this, and in this contribution present a bespoke GPR for extended monitoring active layer dynamics. The work has potential novelty and impact but it is currently in a very early phase and needs further development and analysis.

The authors compare GPR data from a numerical model with experimental data from a laboratory setup. The latter experimental setup for the active layer (described in Section 2.1) is quite comprehensive and appears to be a very good platform to collect measured GPR data. However, only one experimental dataset has been collected which shows some interesting features from the freezing and thawing cycles, but there is no indication of repeatability and sensitivity of experimental parameters. The numerical model is really too simplified and basic to be of much value. It would have been interesting and much more inciteful to include the GPR antenna and dispersive effects from the water (particularly during freezing/thawing) in the numerical model.

The presentation of the manuscript needs to be improved - there is a lack of consistency of style across the figures and the text contains grammatical errors. The features of the measured GPR data (Figure 3c) could be better highlighted and described in the text.