The work is very well presented and of interest for HESS. The statistical approach is robust and the findings of big impact for the hydrological studies of mountain envrionment. The overall process and correlations of SRT with MRS and other hydrological measurements are well developed, and the case studies deserves publication.  The main criticism I have is about the raw input of the seismic data. Authors present just a concise description of the field dataset colection, without showing seismograms or processing phase of the SRT (only some in supplementary material). Authors assert they collect up to 144 channels surveys with 24 channels seismograph (roll?), with 2m spacing (total length up to 286 m, see fig.5), adopting a not clear 8-10 m offset. They used a weak 5kg sledge hammer. The acquisition scheme is not better clarified (roll? Sources? Stacking? Source locations?). By our experience in SRT in mountain slopes, it seems very ambitious to pick first arrivals with such a source over 90-100 m distance. This obviously implies the errors of picking, and then of the inverted section. Authors should provide more information about the raw data collected, presenting clear picked seismograms to prove the timing errors adopted (in paper tab and figures, not in the supplementary materials).

The authors propose a methodology to infer patterns of the subsurface critical zone at the catchment scale from seismic refraction data for hydrological modelling.  The overall study appears to be a very good physically based distributed hydrological model applied to a mountainous catchment. As such, the study is highly relevant and fits within the scope of HESS. 

The manuscript is well structured and well written. Nevertheless, I have some concerns regarding the thickness definition. In the NIHM model, transmissivity (\bar T) is obtained by integrating the hydraulic conductivity between z_b  and z_w in the saturated zone and between z_w and z_s in the unsaturated zone - where z_w is the hydraulic head with respect to the bottom z_b. The water content (\bar θ) is determined by integrating θ in the unsaturated zone, while the storativity (\bar S) is determined by integrating S between z_b and z_w. Thus, by definition, \bar T, \bar θ and \bar S also depend on the state variable, i.e. the hydraulic head.

On the other hand, on page 16 (lines 425-430), the authors state: "The equations defining the groundwater flows show that key hydraulic variables such as the transmissivity \bar T and the water content \bar θ  correspond to the integration over the porous media thickness of the hydraulic parameters K(h) and \theta(h), respectively as stated in (6). Thus, to solve the inverse problem seeking the hydrological model parameters, misestimating the thickness of the hydrological model underground compartments would inherently lead to a wrong assessment of the hydraulic parameters."

In my opinion, the term "thickness" is misleading here. It is not clear whether the author is referring to saprolite or soil thickness, or saturated or unsaturated thickness. In fact, if the former, it is only necessary to rewrite the text, but if they mean the saturated or unsaturated thickness, the dependence of hydraulic head and MRS on both hydraulic parameters and saturated thickness is not new. Please clarify this aspect in the manuscript before proceeding.