The authors have improved the manuscript considerably. However, I still consider that the verification of ET is missing although the authors reiterate that this point is not their objective. It should in my opinion. The main problem with satellite ET products is that these products are never (or at least seldom) verified in the context of the water balance of a basins. I suggest the authors to integrate the signal and create monthly ET climatologies that should be compared against

- over a set o basins. <> means monthly ET over a basin where streamflow has been measured. Q can be obtained from GRDC. P can be obtained from WATCH or Princeton precip products. The results can be presented a in Rakovec et al. 2016 JHM. Such comparison would address one of the critical problems that one of true reviewers pointed out and I fully support. Without verification, what is the point? How should I know which ET should I use?

The Authors have made considerable efforts to address the Reviewers' comments.
I am, however, not satisfied with the final output. Starting with the introduction, authors have put to much emphasis in the study area (The Tibetan Plateau) rather than on the method and research questions. As a results, this manuscript appears to me as a case study rather than as a research article.
I strongly recommend to move the Fig. 1 to data/methods sections and dedicate the introduction to explain what is innovative in this study, which is the key research question and related postulations and how these can be "tested" or falsified within a strict scientific approach. Without this fundamental insight I can not consider a manuscript suitable for publication in HESS.

Authors should expand the method section, results, and discussion.

L. Samaniego