Please see attached review

Dear Authors,

The developed approach is interesting and might be applicable for different landform and climate contexts. The presented work is also a good basis for further studies, which may consider streamflow regimes and shallow groundwater processes to detect headwater streambeds. However, I think the manuscript must be improved prior to its publication, especially regarding to the presentation of results. Please, see below my suggestions and comments:

1. Describe in detail the specific objectives of the study.
2. You should provide some photographs highlighting the main characteristics of the study area as supplementary material.
3. In the text, you mention several times the word “permanent” relating to “stream”. However, I think you mean “perennial”.
4. Table 1: I do not think that roads and urbanized areas have high infiltration rates.
5. Please, provide a flowchart with the methodological steps of the work in the beginning of the methodological section. A short introduction of the applied approach is also valuable.
6. Figure 3: show y-axis in logarithmic scale.
7. You found that PROB is negatively correlated with TPI, with an R of -0.57. Does this multicollinearity have no impact on the presented classification tree models in Fig. 4?
8. Please, improve the presentation of your results, giving more details about them. Moreover, what else can be explored or assessed from the surveyed data? Are there any spatial patterns? What if you compare the results from the different natural provinces?