After reading this manuscript, I question whether the term “catena” is the proper description here. A true catena will have differences in parent materials, vegetation, etc. across the catena. This location appear to have largely uniform parent materials and vegetation. I believe “toposequence” would be a more accurate depiction of the study site than “catena”.

Line 27 – The 5 soil forming factors should not be capitalized.

Lines 74-77 – It is nice that Brantley et al. (2017) developed a conceptual model in three different parent materials, but what is the application/advantage of this model. Why is it important to bring this model up in the introduction? More to the point, what did they conclude as the relationship between reaction fronts and catena position?

Line 109 – The introduction gives a rather abrupt transition from discussing previous works to presenting the objectives of this study. A better transition would be beneficial. And, the objectives should be in a paragraph of their own.

Lines 132-133 – The scientific names of the most common vegetation species should be supplied here.

Table 1 – The latitude in this table isn’t meaningful, because it isn’t different for any of the profiles. Just say the study site is at about 38 degrees N in the site description in materials and methods. Same comment with the downwind bearing.

Figure 2 – It appears that climate and vegetation were varied from 12,000 years BP to present, but were considered constant from 20,000 BP to 12,000 BP. There are not any studies from this area that could provide climate and vegetation data for the 20,000 to 12,000 BP period? This is a definite weakness of the study.

Lines 180-185 – Were any known samples analyzed as a quality check on the data generated?

Table 3 shows chemical composition values for the profiles. However, with the exception of Zr, there is no explanation regarding where this data came from. This needs to be supplied. If the data was generated as a part of this study, how it was generated needs to be explained in the materials and methods. If it was generated as part of a previous published study, that needs to be referenced.

Line 220 – What statistical technique was used to determine that the slope of the line was significantly different than 1?

Table 4 is not needed. Almost all of these values were already given in the manuscript in line 224. Tables should not simply repeat what is in the manuscript (or vice versa).

Line 264 – There are established models with defined topographic positions published in the literature. I do not recognize the model you are using, please provide a reference for it and briefly describe it in the materials and methods. If you are not using an established model, you should either 1) use an established model, or 2) give a complete explanation of your topographic position model in the materials and methods along with an explanation as to why you are using it and not an established, published model.

Lines 318-320 – Is it possible that the deeper weathering profile on the north facing slope is because there is less evapotranspiration, leading to either 1) more water for chemical weathering to take place, 2) more water to allow for vegetative growth, thus affecting chemical weathering, or 3) both 1 and 2?

Line 340 – Are the rainfall amounts in the Betic Cordillera similar to those simulated here? If so, in what way (e.g., are they similar to the mid-level rainfall amounts modelled, the higher rainfall amounts modelled, etc.)? Are there other places that can fill in the missing rainfall amounts (e.g., if the Betic Cordillera amounts are similar to the greatest rainfall levels modelled, are there other studies that can fill in the intermediate rainfall levels modelled)? Right now this modelling in section 3.3 seems weak, in that there is little to no validation. Figure 9 for CDF provides the type of information I would like to see to validate the modeling shown in Figure 8.

The manuscript presents a weathering simulation of soils from Spain. The manuscript's introduction does not place well the current work in the historical...why is this study needed? How does this study fill gaps in the previous work.

Pedon data and profile morphology must be included. In addition, it is not clear what is the genesis of these soils with respect to site geomorphic history, dust or loess deposition. Micro site topography around pedons is not clear. See examples of 9-component slope models, curvature analysis that could enhance your interpretation.

No information is presented on site surface characteristics, percent cover, types of vegetation cover, soil surface integrity, degree of site destabilization due to use (is the area overgrazed?).

To make the case for weathering I believe mineralogy data is an absolute requirement. Simply referring to mafic and concluding weathering when there is a difference in chemistry is not strong enough. I imagine dust influence on these pedons is substantial and CDF might not be the best approach because of this.

More needs to be considered during results interpretation in terms of the effect of site geomorphology and subsurface horizon characteristics.  

I have included an annotated pdf.

Review of the manuscript “Modelling the effect of catena position and hydrology on soil chemical weathering” by García-Gamero et al.

This research work presents application of a 1dimentional pedogenetic (soil profile evolution) model SoilGen on various points on a catena sale landform and comparing the model results to the measured soil properties in the catena. Through this the authors attempt to identify links between topographic position and hydrological attributes with the chemical weathering of soil profiles in said landform. In addition, a sensitivity analysis of the model has been done using a range of annual average precipitation.

From this work the authors have found that the chemical weathering (represented by chemical depletion fraction (CDF)) seems to have no or very weak correlation with catena position or the slope gradient but seems to have some correlation with hydrological factors such as soil moisture and infiltration. The sensitivity analysis of the model has shown some interesting relationship between the annual average precipitation and chemical weathering. Increasing annual average precipitation seems to increase chemical weathering up to a threshold value after which chemical weathering rate seems to decrease.

The manuscript is well written and easy to follow. The results are well presented and the results interpretation and the physical underpinnings of the results are well described. However I have some concerns with the manuscript (described below) that will probably amount to minor revisions. I believe that the manuscript is acceptable for publication after these concerns have been addressed.

1. I believe the readers would benefit form an additional section to the manuscript describing the SoilGen model and its physical underpinnings. The mathematical formulations of such a model maybe too complex to be describe in detail in manuscript like this. However a brief description on how the model works (maybe with a flow diagram) would complete the manuscript.

2. In this work, authors have used slope gradient as the only topographical variable in their analysis. Slope does indeed influence the hydrological state of different areas of the catchment. It is well known that both slope and upstream contributing area (cumulative area of the catchment which drain through a particular point) determine the hydrological state of a node in a catchment. In fact all the landform/soilscape evolution models (SIBERIA, mARM3D, SSSPAM) in 3 dimension use area and slope as primary variables for erosion calculations. As I understand the sample points does not particularly lie on a transect or on the same drainage line. So the upstream contributing area could be very different sample point to point. I believe that an analysis on the relationship of contributing area (or the combination of slope and area) and the CDF could be beneficial to this manuscript.

3. As I understand the slope gradient is kept constant for the entire soil profile simulation time of 20000 years? In 20000 years the geomorphology of the catchment would most definitely change and the slope gradient and (for a lesser extent) contributing area of the sample points would also change. The authors have noted this issue and have attributed the difference in the observed and simulated CDF values for this changes. Were any geomorphological evolution (landform evolution) simulations of the catena done using available landform evolution model such as MILESD, LORICA, or SIBERIA to characterise the change of geomorphological attributes (slope gradient and/or contributing area) of the sample points over the simulation time of 20000years? If the authors have done such simulation work they could use a time series of slope gradients extracted from the landform evolution simulation results at each sample point and input that into their soil profile simulation model which may give better results.

4. Page 9: Figure 2: This figure shows the time series input variables used for the SoilGen model over 20000years of simulations. According to the figure 2 form 12000years to 20000years the input variables seems to be constant. Is this an effect of data not being available beyond 12000years in to the history and average values were used from 12000 to 20000 years or the variables are found to be constant using any other means? This was not clear from the text describing the figure2.

5. Page 18: Figure 8: In this figure the authors have provided the results of the sensitivity analysis of the model with respect to different annual average precipitations. The CDF of the upper soil layers (<40cm) increase with precipitation until 800mm/year and then decrease. The general pattern of the CDF-soil depth variation is consistent for all the CDF-depth curves where the CDF is high in shallow soil compared to deeper soil, except for the 900mm curve where the CDF of the upper soil layers are lower than the deep soil layers. The 1000mm curve seems to show the same trend as well. It would be beneficial to the readers if the authors could elaborate on this “inversion” of depth vs chemical withering curve at 900mm and the physical underpinnings they suspect leading to this anomaly.