The authors quantified evaporation/sublimation (E) during ice-free (IF) and ice-cover (IC) periods for a large lake on the Tibetan Plateau. Field observations were collected between 2014 to 2019 and used to quantify evaporation/sublimation (E) and determine the main controls on E during the IF and IC period and annually. These results were then used to validate and assess three different types of E models (Mass Transfer, atmosphere dynamics and statistical model) to determine which model(s) would be adequate for simulating E during IF, IC and Annual (AN) conditions. The models were introduced to simulate E for the 2003 to 2017 period using reanalysis data to study climate change during IF, IC and annual lake conditions. This paper presents an interesting and innovative contribution to lake E by using 6 years of continuous high-resolution and precious observation datasets. There are not too much papers assessing evaporation from the Tibetan Plateau region or studying sublimation during the ice-covered period. The significance of the results is thus important for improving our understanding of the main controls of E during both IC and IF conditions on an alpine saline lake, and these results can be helpful to improve current hydrological models of alpine lakes. Thus, I recommend this paper for publication in HESS after a major revision. Besides, I did have some concerns about this paper as follows:

Major comments:

(1) The objectives contradict some of the methods. In the second objective, the authors state that two models will be calibrated and verified, however, within the methods section three models are calibrated and verified and not just two models.

(2) Use summary tables for the observed data collection, Reanalysis of datasets, models, and variables. This will make it easier to understand the data collection, cleaning, and processing. Currently, the way these variables and their measurements are presented makes it unclear. For example, in Line 138 it is not clear if the gas analyzer is at the same height as the 3-D sonic anemometer. Besides, the observed meteorological data is in a 30 min timestep; but ERA-5 is in a 1-hr timestep. How was this addressed when assessing the fit between the observed data and the reanalysis data?                                                                                                                                                                                                                                      (3) E values for Antarctica are in mm/month during IC, Lines 346-347 you present the annual sum of E; but to draw comparisons to Antarctica can you put this value into monthly for the IC period? The total value does show it is larger but by showing it in the same units as Antarctica it will be easier to see how it relates monthly

(4) In the key findings you state that wind weakening is considered a key finding; however, wind weakening and its relationship to E during the IC period is not discussed. As this is considered a key finding this should be discussed.

Minor comments:

(1) Line 37: did the result for IC consider ice loss?

(2) Line 132: you should reference your site in Figure 1.

(3) Line 166: Long time should be long-time.

(4) Lines 178-183: Qui et al 2019 is the referenced method for the ice phenology dataset, however, how do they account for the accuracy of the ice dataset you are using for your analysis? Using visible MODIS to ascertain freeze dates can be difficult, as the ice must be substantial enough to change the reflective properties. A few brief sentences to expand on the methods in this section would do well to provide context for the accuracy of the ice dataset you are using.

(5) Fig S3: the x-axis should be the same for all 3 figures. They should all range from 0 to 60%; if you are to just glance at the figures and not read the axis label/units one would assume they all contribute the same during each period.

(6) Fig S5: the y-axis should have the same scale for all figures. Why is the x-axis for ice cover 1 year? Whereas the IF and AN showing 3 and 4 years respectively? Your caption states they are showing the results from 2014-2018.

(7) Fig 1: DEM needs units, missing the line for rivers in the legend, is the scale the same for the inset map?

(8) When using the abbreviations for ice-covered (IC) or ice-free (IF), they are missing context (or a word) such as conditions or periods.

Evaporation measurement and modelling of an alpine saline lake influenced by freeze–thaw on the 2 Qinghai–Tibet Plateau

The paper is focused in exploring the key factor and salinity on E. The microclimate factors are well explored buy in regard the salinity with some weakness for two different conditions considered (IF, IC). In general, the paper provides an important technique contribution.      

59-61 – The paragraph includes the key and rich knowledge on saline lakes E, but is included only one author, Hamdani et al., 2018.

117 – In this line is referred to climate changes for the period 2003-2017, why? Maybe is it correct to refer only to climate variability as it written in the abstract.

121 – In title 2.1 would be better split into two chapters

In my opinion the Site description is poorly described, how about the other lake characteristics such as: lake topography, inflow-outflow, stratification, thermal stability, hydrodynamics?

128 – The sentence seems nonconclusive, it does not include any information and Reference.

151 – Why the water temperature (Tl) was measured only at depth till of 3.0 m?

264 – To explore the key factor controlling E, was used two methods to estimate the sensitivity and the importance of each variable. Can you define the difference between the two approaches, or add more literature?    

362-385 - Was discussed the influence of salinity on E rate. Then E-IF and E-IC rate within the saline lake environment in consequence is one of the interests. I think the literature explains quite enough the salinity influence on E, but this paper attends to describes the within the two different thermal conditions and no results. Is explained that was measured the water activity by 0.97 and applied to the model, that is it (IF, IC ?).

In the same long paragraph, is defined the reduced saturated vapor pressure above the water (at a given water temperature, which one?); in other wise, there must be more evaporation, but here is given the opposite definition.

This section must be clarified.    

Dear Editor,

Can you provide the list of paper:

Hamdani et al., 2018

Salhotra et al., 1987

Wang et al., 2019ª