The authors present a wavelet-based framework for understanding the variability in precipitation, temperature and streamflow in France. Especailly interesting is the cross-scale analysis. However, the discussion of these results were not well done. The authors have only reported the results. More discussion on these trowing light on the physical intuition and learning derived is needed. It was also not clear if the cross scale implied across spatial scales ... in which case, one driving others is confusing as the manifestation  of a variable, while having spatial variability, need not necessarily always have a causal connection. In the case of streaflow, this is clear due to the network nature of the flow. If the cross-scale is across wavelet scales, in that, for each time series, the dependence is between low-frequency vs. high frequency, then how does this dependence manifest in space needs to be disscussed. 

The paper by Fossa et al. describes spatial interactions and cross time-scale features of variability in hydro-climate variables across France. The work is timely and of potential impact, it is relatively well-written and the cross-scale analysis is quite innovative. Therefore, I see potential for publication in HESS, but only after the manuscript will undergo a substantial revision.

In fact, two main points must be addressed with care before the paper can be accepted:

- the methodology cannot be taken for granted for the majority of the readers of this journal. Figure 2 should be a milestone of this manuscript, while instead it is useless in this form, and likely wrong since the caption does not correspond to what is plotted. Instead, figure 2 should clearly show the workflow that leads the reader to acknowledge the results. Also, more explanation on the meaning of the wavelet significance test should be documented.

- the results are not described in depth, and also they lack of a discussion that is necessary to enrich the paper. In particular, a few figures are almost neglected in the description, but in general all results are not discussed in depth, rather there is only a documentation of what can be read in the plots.

Specific comments can be found in the attached documents (I recommend to open the document with Acrobat Reader).