Comments on the ms “Long term variations of actual evapotranspiration 

over the Tibetan Plateau” by Han Et al.

General comments

The authors did an interesting and challenging research, as it is known that the ET is essential to water and energy cycle on the TP. The authors estimated the ET from 2001 to 2018 by using SEBS model, which has contribution to understand the water cycle on the TP, while the MODIS products should be used carefully over the TP for its complicated weather and underlying surface. The authors, please, add more information in detail of the MODIS data used in this study.

And, in this investigation, the estimated ETa by using the SEBS model were validated with six flux tower data from EC observation. It could show the estimated ETa was reasonable used over the whole TP, however, the results are not fully convinced, and the Rn , air temperature and velocity should be evaluated by using the observation data at 6 sites, and then extend to analyze the variations of ET in the western TP ,Eastern TP and the whole TP.

In general, I would like to recommend accept this manuscript after minor revision and publish it in this journal.

Specific comments 

P2: line 37-38,” The domain mean of annual ETa  on the TP decreased slightly …”should give a reference. 

P3: line 57-58, the sentence does not make sense, please make it clearly.

P4:line 78, ET and its variations have been drawing more attention worldwide. 

P5: line 95-102, the authors listed three studies, how does their results performance? Do the authors compare the results (Ma, 2019) with the 6 flux tower observations? It is not clear why the authors conducted this research, just like the authors mentioned, Ma, et al, 2019 they got the Eta from 1982-2012. We know the two methods are both used to estimate ETa at regional and global scale.

P7: line148-149, the “the”, in the net radiation flux, the latent heat flux, the sensible heat flux, the ground heat flux, should be removed.

P7 and P8, line 169-196, whether the parameter, d0,Cp  and U * are also from CMFA?

P11, line 272- 273, the specific data should be used to show how does the SEBS performance well at the two sites. 

P25/P27: The length-width ratio of Figure 1 and other Figures is different.

Long term variations of actual evapotranspiration over the Tibetan Plateau 

C.B. Han, Y.M. Ma, B.B. Wang, L. Zhong, W.Q. Ma, X.L. Chen and Z.B. Su

This study describes a data set on estimates of monthly evapotranspiration on the Tibetan Plateau for 2001-2018. The estimates are based on SEBS and a diverse input data including both satellite retrievals and re-analysis data. The estimates have been evaluated against measurements acquired by six eddy-covariance systems located on the Tibetan Plateau. The dataset is freely available

General comments

The Authors have to be commended for the effort to generate, evaluate against in-situ measurements and make available a potentially useful data set on ET. On the other hand the manuscript has a number of shortcomings, detailed below, which should be addressed prior to publication.

The major issues are:

1)    The entire study and the applied model as described seem to assume that the only relevant water phase transition is liquid to vapour, either as evaporation or transpiration. Clearly, on the TP all water phase transitions occur at different times and place and should be taken into account to produce monthly estimates of ET, or better LE. Additional comments are given below.

2)    Temporal sampling of the data set is not explained and, as a matter of fact, model description does not even mention any temporal dimension. How are monthly estimates actually obtained?

3)    There are multiple instances of poorly explained land surface processes related to the water phase transition.

Specific comments

L29 The Abstract should list input data applied.

L71 Not clear whether it is really meant that ET supplies 2/3 of P, in which case the obvious question would be where does the remaining 1/3 come from. This sentence may also mean something completely different, i.e. that land ET is 2/3 of P overland, implying that the remaining 1/3 may come from oceans. Please clarify

L75 modulating in which sense?

L76 “only connecting component” not correct: all water phase changes involve large energy exchanges.

L93 “according to contrasting trends 93 between Epan and actual ET…” This is all very confusing, since Epan and ETa cannot be compared. In addition Epan is largely affected by the configuration of the pan. There is an old FAO Technical Bulletin dedicated to the measurement of pan evaporation, which describes in detail how the design of a pan affects the measurement.

L94 this sentence is misleading, since it suggests that Epan may provide information on Eta

L107 “plateau-scale variations of ETa” Not quite sure I understand this sentence. To measure spatial variability and patterns a number of measurements at different locations are needed, so what is wrong with a limited footprint of each measurement.?

L121 Model description does not include any information about parameterizations applying to snow and ice.

L143 “2.1 Model description” On the TP snow and ice cover a large area, with snow cover varying rather rapidly in time. There is not a single comment about this and SEBS as described does not account for energy and mass exchanges between snow / ice and the atmospheric boundary layer. Moreover, the transition liquid to vapour is not the only one determining the water mass and surface energy balance. 

L145 “remote-sensed land surface energy” at which time interval? no mention in this entire Section of temporal coverage and sampling.

L149 “latent heat flux” L is different for melt, sublimation and of opposite sign for condensation, freezing and deposition. Is this taken into account?

L164 “Over water surfaces (NDVI < 0 and a < 0.47)” Implications of this sentence not clear

L166 “G0 is negligible” This is a very peculiar statement, since it is the heat absorbed by a glacier that drives melt. Also, the thermal conditions of glaciers are far from stable, even though temperature may remain < 0.

L187 Eq.7 is h the same as z in the previous equations?

L190 rather unlikely that a DTM can capture roughness elements like rocks and similar details.

L198 “an evaporative fraction” What is this supposed to mean? there will always be an evaporative fraction, regardless of whether a model is applied to estimate it and even less relevant is which model is applied.

L199 These sentences denote a poor understanding of the fundamentals of energy balance at the limiting conditions.

L202 “evaporative fraction” The evaporative fraction is not the residual of the surface energy budget.

L207 “CMFD” add full denomination the first time used

L211 which data set?

L236 “post-processing” of what?

L240 “the energy

L241 “storage in the layer above” In snow and ice the latent heat of fusion should be added

L264 “phase” I guess this depends more on whether the sampling interval is short enough to capture significant fluctuations in the ET signal. Do not see how this might be inherently related to a model.

L282 “mean” does this mean the average of annual ET between 2001 and 2018?

L292 “have less available energy to evaporate.” But snow and ice will melt....

L295 “evapotranspirated” even in case the approach described would be applicable to any surface including snow and ice, the estimated LE would relate to the net latent heat balance. i.e. to the net energy absorbed and released by all water phase changes, not just to evaporation and transpiration.

L303 “Note that the distribution pattern almost faded out in winter season” faded out in which sense?

L349 define amplitude and changing rates, otherwise rather ambiguous in this section on trends.

L356 “decrease of Rn” This deserves more attention, since it might be related to increasing albedo as due e.g. to increasing snow cover.

L366 “the melting of permafrost and glaciers on the TP. Hence, the melting water.” no attention paid, in Model description or anywhere else, to the associated latent heat of melting!

L374 “high-spatial resolution” which data sets? is MODIS high resolution? any consideration about length-scale of spatial variability? At L378 10 km is indicated as spatial resolution of the ET estimates, this is by n no means high spatial resolution, especially given the spatial variability of snow and ice.

L389 “evapotranspiration” This is a very misleading term to describe all the water phase changes on the TP.

L407 “rate of change” is this different from the trend mentioned elsewhere?

L431 “forest canopy height” where used? Model description does not mention canopy height as a relevant variable and does not explain how such data might have been used.