The reviewers have found the goals of the manuscript to be quite interesting, but have also identified serious shortcomings. Major revisions that address all the points made in the detailed review comments will be needed to make this paper suitable for publication. Of particular importance will be moving beyond a fitting of the data to an actual test of the predictions.

The second round of review came back with widely divergent referee judgments. My own judgment is that the paper is an important one, and that it has improved considerably in this revision. I also believe that in the commented version that is attached I was able to address the language issues. The positive reviewer has, nonetheless, come with a number of constructive suggestions that would further improve the manuscript. The more negative reviewer also provided several constructive comments. I believe that addressin all of these issues are relatively straightforward to implement (especially relative to the changes made in the last revision). I am therefore returning the manuscript with an instructions for further revisions noted below. If you feel that you are able to make these changes, I will look at the revised version and make my editorial decision without further referee review.

Comments to address from the reviewer who recommended rejection:

The validation of the proposed approach (included as per both reviewers' suggestions) show that the proposed approach does not provide satisfactory results in 3 out of 7 ecosystems (see Fig. 8).
The results still denote a lack of understanding of the mechanisms driving evapotranspiration in different ecosystems – e.g., the lack of leaves or even plants (for crops) during the winter months. Despite some suggestions on how to handle this issue in the previous round of reviews, the author did not address this point. Rather, in the result section, the authors discuss patterns of average annual Kc, which is not very meaningful when considering temperate and boreal sites (as apparent from Fig.1). Further, they recognize that the seasonal pattern of Kc in evergreen vegetation is more stable in deciduous ecosystems, yet they fail to state why this is the case (needles are retained throughout the year, as opposed to the situation for deciduous trees).

Second, the validation now introduced as part of the revisions, clearly shows that in several ecosystems the approach does not work very well (Fig. 8). The modelled vs. simulated evapotranspiration rates, while correlated, do not fall on the 1:1 line (or at least near it) in at least in 3 out of the 7 ecosystems, leading to overestimated or underestimated values. Interestingly, one of the ecosystems in which the approach is not working is the crops, where the FAO model was developed.

Third, there are several unclear or incorrect claims. I report here some examples:
L 64 How can PET be considered stable when (the author acknowledge) it depends on temperature and precipitation?
L 90 The FAO approach has been used for many more crops (and the Kc values are tabulated in Allen 1998, for each and every growth stage).
L 129 How were the ‘validation sites’ selected? Where are they located?
Fig. 5: I am confused by the evergreen broadleaf forest at 60 deg N (one of the two sites among the evergreen broadleaf forested ecosystems). Probably it would be worth to provide a bit more information about the sites, particularly when very few (and hence potentially non-representative) are available

Comments to address from the reviewer who recommended publication ( I am not sure the authors got to see this as the reviewer submitted these confidents in a channel that may not have been available to the authors):

The article of Liu et al. is a needed contribution to studies of evapotranspiration rates across ecosystems and regions. The methodology is consequent and clear. I enjoyed reading the article. The multi-linear models developed by the authors for the ratio of actual evapotranspiration to potential evapotranspiration will be an important tool for hydrologists on the field.
However, some aspects need to be improved. Some key studies are completely missing from their manuscript. Starting by the studies of Budyko (1974) were potential and actual evapotranspiration are put in context. I suggest other important references that could better support the discussion.

I suggest an inclusion of a Figure that compares water and energy use efficiency for all the ecosystems compared by the authors. This comparison would enable a direct comparison of Kc (AET/PET) and evaporative ratio (AET/P) for all the ecosystems evaluated in the manuscript, enriching the discussion. See Van der Velde et al. (2013).

Also, general information on the FLUXNET measurements should be included in the manuscript. See below some typical questions.

Any autocorrelation between latitude, precipitation, LAI if not treated independently? I would say that in these Northern Eurasian latitudes as you move northwards you get more rain?

Other aspects along the manuscript are found in detail below:

First paragraph and Line 59-You could mention briefly here that the uncertainty in AET is mainly due to all the factors affecting vegetation AET rates as mentioned by Jaramillo et al., Journal of Hydrology (2013) or Donohue et al. Journal of Hydrology (2007), Hasper et al., Functional Ecology (2015) and by all the climatic and landscape drivers of ET change (See Jaramillo and Destouni, GRL, 2014). Even better, mention the most important.

Line 86-89- Shown by Zhang et al. (2001)

Line 122- Upfront, please specify the time period of ET availability from FLUXNET, time-scale, how it was obtained, etc. A brief summary could be useful for the general reader that does not know of eddy-flux ET measurements.

Line 134- in what units are AET and VDP being measured by eddy-flux towers?

Line 139-How did you estimate all these parameters, Rn, slope of sat, etc? What assumptions of Allen et al. 1998 did you apply? Just explain what assumptions did you use, mention the equation numbers in Allen et al. 1998. What height was wind speed measured at in the towers? With what time scale was PET estimated? Was it later aggregated in time to agree with the AET time scale?

Line 163-Can you send to Supplementary a figure showing how AET and PET vary from month to month. This would enrich the discussion to understand the variations of Kc from month to month!

Ecosystem Acronyms-Can you mention again in the results what each of these acronyms are, EBF, GRA, etc, it is difficult to go back to the first explanation every time you use one. Or maybe use easier-to-understand abbreviations.
Line 174-What do you mean with this?

Line 217- by ecosystem, Line 220- measurements with

Line 240-I think this should be stated right from the beginning, that some of the flux measurement sites are in irrigated areas! You know, irrigation has been proven to be driving ET changes at the local and even at the global scale. See Jaramillo and Destouni, Science, 2015.

So when plotting the figure I just suggest, irrigation could represent much of the high AET or Kc rates. See Van der Velde et al. (2013).

Are there other sites that have irrigation in your study? The irrigation issue should be mentioned in the FLUXNET methods.

Line 244-You should say that this mainly occurs since as latitude is decreasing, PET increases, but AET increases even more than PET. It is the only way this can happen. This is an interesting finding.

Line 248- Mention this in the beginning of the FLUXNET methods.

Line 260- Sorry to be pushy, but again a more updated study showing the domination role of irrigation on ET such as Jaramillo, F., Destouni, G., 2015.

Line 266- connect the two sentences

Line 284- What is leaf resistance?

Table 1- What do blank spaces mean? Non-significant values?

Figure 1- Legend is missing

Figure 3- Explain uncertainty bars, are they one std dev?

Conclusions- the conclusions should state that the models apply to northern temperate and boreal latitudes, and that its extrapolation to other tropical and southern latitudes should be explored.

Some suggested references that could enrich the literature review and discussion mentioned in this review

-Budyko, 1974. Climate and life. Academic Press.
-Donohue, R.J., Roderick, M.L., McVicar, T.R., 2007. On the importance of including vegetation dynamics in Budyko’s hydrological model. Hydrol. Earth Syst. Sci. 11, 983–995.
-Hasper, T.B., Wallin, G., Lamba, S., Hall, M., Jaramillo, F., Laudon, H., Linder, S., Medhurst, J.L., Räntfors, M., Sigurdsson, B.D., Uddling, J., 2015. Water use by Swedish boreal forests in a changing climate. Funct. Ecol. n/a-n/a. doi:10.1111/1365-2435.12546
-Jaramillo, F., Destouni, G., 2015. Local flow regulation and irrigation raise global human water consumption and footprint. Science 350, 1248–1251. doi:10.1126/science.aad1010
-Jaramillo, F., Destouni, G., 2014. Developing water change spectra and distinguishing change drivers worldwide. Geophys. Res. Lett. 41, 8377–8386. doi:10.1002/2014GL061848
-Jaramillo, F., Prieto, C., Lyon, S.W., Destouni, G., 2013. Multimethod assessment of evapotranspiration shifts due to non-irrigated agricultural development in Sweden. J. Hydrol. 484, 55–62. doi:10.1016/j.jhydrol.2013.01.010
-van der Velde, Y., Lyon, S.W., Destouni, G., 2013. Data-driven regionalization of river discharges and emergent land cover–evapotranspiration relationships across Sweden. J. Geophys. Res. Atmospheres 118, 2576–2587. doi:10.1002/jgrd.50224

I would like to commend the authors on this ambitious manuscript, and the willingness to work with the reviewers on revising it.