Contrasting hydrological and thermal intensities determine seasonal lake-level variations – a case study at Paiku Co on the southern Tibetan Plateau

Lei, Yanbin; Yao, Tandong; Yang, Kun; Lazhu,; Ma, Yaoming; Bird, Broxton W.

doi:https://doi.org/10.5194/hess-25-3163-2021

Articles | Volume 25, issue 6

https://doi.org/10.5194/hess-25-3163-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-25-3163-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 25, issue 6

Research article

|

10 Jun 2021

Research article |

| 10 Jun 2021

Contrasting hydrological and thermal intensities determine seasonal lake-level variations – a case study at Paiku Co on the southern Tibetan Plateau

Yanbin Lei, Tandong Yao, Kun Yang, Lazhu, Yaoming Ma, and Broxton W. Bird

Download

Final revised paper (published on 10 Jun 2021)
Supplement to the final revised paper
Preprint (discussion started on 13 Jul 2020)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Comments to “Thermal regime, energy budget and lake evaporation at Paiku Co, a deep alpine lake in the central Himalayas”　by Lei et al.', Anonymous Referee #1, 20 Jul 2020
- AC3: 'reply to reviewer #1', Yanbin Lei, 14 Sep 2020
AC1: 'Reply to reviewer 1#', Yanbin Lei, 29 Jul 2020
RC2: 'Comments', Anonymous Referee #2, 05 Aug 2020
- AC4: 'reply to reviewer #2', Yanbin Lei, 14 Sep 2020
RC3: 'Review of Hess-2020-320', Anonymous Referee #3, 18 Aug 2020
- AC2: 'Reply to reviewer 3#', Yanbin Lei, 06 Sep 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Publish subject to revisions (further review by editor and referees) (02 Oct 2020) by Bob Su

AR by Yanbin Lei on behalf of the Authors (13 Nov 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (24 Nov 2020) by Bob Su

RR by Anonymous Referee #1 (02 Dec 2020)

Suggestions for revision or reasons for rejection

Comments to “Contrasting hydrological and thermal intensities determine seasonal lake level variations A case study at Paiku Co in the central Himalayas” by Let et al.

Contents are improved to article level showing important consequences based on the observation. This paper focused on the mechanism of high lake evaporation to cause rapid lake level decrease in post monsoon that may be due to climate condition around Paiku Co. This evidence is valuable to be published in HESS if the following issues are re-considered.

Major comment
1) As section 3.4 with Fig. 6 would be the most important parts, in-depth explanation with discussions are expected with re-structure chapters. If you want to divide Result and Discussion, the Discussion part should contain comprehensive issues mentioned in the Result. As the 4.1 treats some kind of execution for observational accuracy, beside 4.2 compared with other lakes overlapping to 3.4, I would like to recommend, delete “4 Discussion”, “5 Conclusion” may change to “5 Summary and discussion” if possible where the problems or future challenges such as 4.1 should be covered there, and 4.2 should move to Chapter. 3. (if the article HESS format accept.)
2) Data and units are very important for papers based on the observational confidence, and many of them are missing. There are also uncertainties in the label and amount calculation. For instance, Fig.4, 5 are all daily value? What is the original observation interval to calculate heat/water budget? How did you filter errors/missing with correction or interpolation to born Fig. 4-6 ? In Fig.6, is this evaporation/precipitation rate? If so, they should be mm/???.
3) Total river discharge was about 800-900 mm (per what??), and Av was not considered (L128). On the other hand, at L226, total evaporation for M-D was 975 mm and that is a matter of concern. They are the same order. By means of heat budget, yes that heat advection and heat needed for evaporation/condensation is quite different. But, still I can not understand the logic explaining the lake water level changes only by the evaporation term (river discharge data is missing in post-monsoon at Fig. 6b). Problem would be that there is no precise explanation about the water budget depending on the season including precipitation amount at around L125. Also, G is ignored (L128) but calculated at L151. Very strange.
4) Seasonal variation of wind speed, as shown in the reply sheet, is valuable to explain the characteristics of lake temperature stratification. Why don’t you add the wind speed variation on the Fig. 2 ?

Other comments
L16　lake evaporation was “estimated” ?
L18 results showed ,, (results should be passed sentence)
L23 It is better to show lake evaporation amount for pre-monsoon, monsoon, and pos-monsoon, respectively, to emphasis their seasonality.
L25 Better to unify “summer monsoon” and “monsoon”.
L27 “shallow lake” ,, again, please explain which level is shallow and which level is deep ? Also, lake depth should be added at L49,51.
L83 Paiku Co tends to lower the lake level. So, how this evidence relates to this study’s results? Better to be discussed in the last chapter.
L119 Just my impression that G should be ΔS　and S should be G? Because, G came from Ground, and S is Storage.
Fig. 4 Spell out SVD,LWD,LWU
Please check the order of reference again.

Hide

RR by Anonymous Referee #2 (20 Dec 2020)

RR by Anonymous Referee #3 (24 Dec 2020)

Suggestions for revision or reasons for rejection

General comment:
The authors revised their manuscript and now it is in better shape. However, the main issue which I pointed out in my previous review (and in the reviews to the previous version of their manuscript) has been solved only partially. They did not explain why in their response, either. Since this is the second round of the review process (and fourth if we count it from the previous version), I recommend "reject" of this manuscript in its current version. The details are explained below.

Main issue:
It has been the accuracy of the estimated evaporation in their study. They used on-the-shore measurements of temperature and humidity with an instrument designed for indoor use without ventilation/radiation shield (except for an instrument container box) to estimate the above-the-lake Bowen ratio. Since this is not where and how they should be measured, it is the responsibility of the authors that their measurements produce evapotranspiration with sufficient accuracy to allow them to claim their findings are meaningful against errors. My suggestion of the previous review is (1) to identify all the error sources and estimate each error magnitude, and then (2) to estimate error propagation to the Bowen ratio and latent/sensible and fluxes. Authors did (1) but did not go on to carry out (2). They do not explain why they did not in their responses (except perhaps in L284 saying "the difference of 0.6°C .....is used to estimate the uncertainty of lake evaporation" but this estimation is not shown in the main text. Authors should respond (item-by-item) to every comment made by reviewers. If they do not agree, that is fine. But they need to explain why with supporting evidence. Please also note that some of the error estimates are for the whole year. But what is really needed is the seasonal changes of errors since they focus on seasonal changes. We need to know if the magnitude of the seasonal variation of evaporation shown by the authors is larger than the error bars. Note also that they do give their error estimate of evapotranspiration based on water balance. However, this is a rather crude estimate (see minor comment on L302-310 given below) and should be supplemented with the micrometeorological error estimates.

Specific comments:
Title: "in the central Himalayas"; in their manuscript, they mainly explain lake evaporation and water level in the Tibetan Plateau, and not in the central Himalayas. Since not all HESS readers are familiar with the geography over there, authors should mention the relation between the two, or they should only use the Tibetan Plateau throughout the manuscript.

L53-54 "The main causes ....still unclear"; aren't there any theories or hypotheses presented in previous studies? Since this will be discussed as the main point of this study in section 4.2, previously proposed ideas should be thoroughly reviewed and explained here.

L56, L321 "direct measurement... energy budget method"; in my previous review, I made a comment on this saying that Bowen ratio or energy balance method is not generally considered a direct method. In the response of the authors, they simply say, "We have deleted this paragraph."(their response, L333). I take it they have agreed. Then they should revise the same point throughout their manuscript. If they do not agree, they should explain why.

L63-66 "...not typically investigated through eddy covariance system because....difficult ... measurement platform"; if it is the platform issue, then the same applies to the Bowen ratio system, doesn't it?

L70 "ice-free period"; they do report evaporation and latent heat flux during the ice-covered period in Figs.5-6.

L82 "highest shoreline"; is this paleo-shoreline?

L93 "water temperature profiles were"; add sensors after profile to write "water temperature profile sensors were".

L96 "daily-averaged values"; I assume they are used only for thermal regime analysis of the lake. For the application of the energy balance method, weekly-averaged values are used, aren't they? Please clarify in the main text.

L100 "over the lake"; this does not express the exact location of their measurement. It should be something like "above the shoreline".

L103 "accuracy"; mention that it is the accuracy of sensors without considering the error arising from lack of ventilation and a radiation shield.

L103 "Fig.2" this is the wrong figure number.

L128 "very small and can be neglected"; they make this statement based on the estimated annual lake water temperature decrease (about 0.1°C). However, it should be based on the estimation of seasonal impact on the storage in the energy unit. The impact should be different depending on the season (e.g., the 2-4°C higher water flows into the lake from July to December), and even a small temperature change could produce a large storage change since water volume is quite large.

L128 "we do not consider...G"; they do consider G with Eq. (5). Perhaps this is precipitation? But then no discussion is made on the impact of rainfall advection. It should be made if G should be replaced with precipitation. If G represents "the heat transfer between the lake water and bottom sediment" (as mentioned in their response, L366-367), they still need to explain in the text why it was ignored. Note that the reason "because there is no data available" (their response, L366-367) cannot be a valid reason by itself; the reason has to be, first, scientific (and then perhaps practical).

L134 "Ra"; is this Rl?

L141-142 "the two uncertainties...... cancel each other ...daily"; I do not think this is always true, and whether or not they cancel out depends on the specific condition of the diurnal changes of wind speed, solar radiation, etc. For example, Wilson et al. (2013, doi:10.1002/jgrd.50786) and Sugita et al. (2020, doi:10.1029/2020WR027173) report the cases where daytime and nighttime values do not cancel out. I am sure some reports indicate mutual cancelation. It is condition-specific.

L151 Eq. (5); in this equation, the summation is from i=0 to 72.8. This should be from 0 to 13? What would you do when the water level changes? Layer number remains the same, but the representative depth of each level changes?

L166 "reduction in wind speed (data not shown)" and L175 "significant increased wind speed.(data not shown)"; I commented on this in the previous review report. The authors' response explains that they obtained this result of "reduction" based on the Qomolangma station measurements (their response, L377-380). Then they need to explain this fact in the main text, perhaps in section 2.2 around L112. Also, they need to explain how similar the wind speeds are between the Qomolangma station and Paiku Co to make their statement valid. By the way, wind direction should also be an important factor, particularly for the estimation of measurement errors with the sensors on the shore. When wind direction is from the lake to the shoreline, this is a better situation while the opposite condition would produce a larger error.

L186 "lakes on the TP are considerably lower than those in other parts of the world (Livingstone,......)"; two references are not available in the reference list, so I cannot check what "other parts of the world" means. Valid comparisons should perhaps be made between lakes in the TP with lakes of similar size and depth, and in the same latitude but in lower altitude.

L187 "due to the low lake surface temperature"; please explain the logic here. I would think that when summer surface temperature is low, the temperature in a deeper layer should also become low. Therefore gradient may not be affected much by this fact.

L190 "3.2 Energy budget"; with this title, the sensible and latent heat fluxes should be treated here. Currently, it is in "3.3 Lake evaporation" section.

L196-198; are these numbers in the parentheses the averages during the specified period? Please indicate what it is in the main text.

L222 "during the ice-free season is shown in Fig.6b"; Fig.6(a), rather than 6(b), does show evaporation during the lake ice periods.

L226-228 "lake evaporation ....not determined because ....covered by the lake ice..."; since it is reported in Fig.6(a), why not? The presentation has to be consistent throughout the manuscript.

L240 "Fig. 6b,c)"; they should be 6a and 6b?

L290-291 "MODIS derived lake surface temperature"; add information/reference of this product. Is this an instantaneous value? Or is it some kind of time averages? How many data points were used to derive the difference? And spatial resolution? Is it small enough to match with in-situ measurements, without the land surfaces in the field-of-view?

L295 "whole year"; estimation should be made in shorter time intervals to allow seasonal changes of errors of the Bowen ratio/evaporation (see major comment).

L302-310; is it correct that mean evaporation and mean lake level decrease were determined for May (one month) and Oct-Dec (3 months)? Is it also correct that river runoff contribution was determined once in spring and once in Autumn for three years (so the total number of measurements are 6)? Is it reasonable to compare the 6-day measurements with one- to three-month averages? Please clarify and explain them in the main text. Then why only October discharge contribution was considered? Why not the spring contribution? Is the stated contribution of 1.2 mm/d (L307) correct? A simple average would produce 1.0 mm/d. Also when you use water level, you need to consider the error magnitude of the water storage estimates (L300).

L303 "Runoff measurements"; add an explanation on how, when, and at what intervals the measurements were made in the method section.

L311 "4.2 Comparison with other lakes on the TP"; authors discuss the differences between deep and shallow lakes here. This is the issue explained in the introduction section as "unclear" so that it is good. Then how about another issue (difference of the range of seasonal lake level variation between southern and northern Tibetan Plateau) mentioned in the introduction section L45-46. Any new findings on that in this study?

Table 1; the listed values of the accuracy of GMX600 are not accuracy but resolution. I cannot find the accuracy of "5%" in the specification of CNR4.

Fig.6; change the unit of mm in the y-axis title to mm/month, mm/week, or mm/day.

Fig.7; add units to the y-axis. There are two grey rectangular shapes; add an explanation to each.

Authors' response L348-350 "daily Bowen ratio is calculated...."; this should be explained in the main text around L145-150. As it is, the description is "all the input data were averaged at the weekly interval before lake evaporation was calculated" (L149-150), and therefore we come to believe that the Bowen ratio was calculated from weekly average temperature and humidity.

Authors' response L407 "... heat advection due to precipitation... has been mentioned in the revision"; in the main text, it simply says "the influence of ...precipitation is very small and can be neglected" (L127-128) without supporting evidence. Add evidence, please.

Authors' response L530 "We have addressed... in Fig.8...."; I do not see it.

Hide

ED: Publish subject to revisions (further review by editor and referees) (29 Dec 2020) by Bob Su

AR by Yanbin Lei on behalf of the Authors (31 Jan 2021) Author's response Manuscript

ED: Referee Nomination & Report Request started (14 Feb 2021) by Bob Su

RR by Anonymous Referee #1 (26 Feb 2021)

Suggestions for revision or reasons for rejection

Most of the explanation to reviewers’ comments were done in the contents, but they are still scattering and additionally. I would like to suggest brush up the story, especially for following matters.

Major comments
1) According to the abstract, one of highlights of this study is that evaporations from the lake were estimated with their differences depending on pre-monsoon, monsoon, and post-monsoon seasons. However, there is no clear explanation that why the defined months correspond to the actual seasonality of monsoon at Paiku Co. Monsoon variability can be identified by precipitation or wind condition at the lake, however, there is no analysis about the monsoon itself. Please add the explanation about in-situ monsoon seasonal changes to match the months. Also, “summer” is used in chapter 3.1 and “Spring” and “Autumn” are used in Table 3. Please unify the term of season.
2) My understanding is that heat budget at the lake surface was estimated by Bowen-ratio method where the ratio was determined by the temperature and humidity profile at the lake shore. But, downward radiation components were observed at another station 150 km far, and albedo was also used as constant value (0.07) based on a reference (I could not find it in the reference, but it is not about the lake on TP). Upward L was estimated by the lake water, not the skin water temperature. Author need to explain those key concepts clearly at first in Chapter 2.3. Observing net radiation (Rn) is quite important for this method, and also important for results because comparison between the evaporation and Rn was discussed. I wonder the heat budget can be closed, if you use downward radiation components in such far station (Fig.S4 may help?) and also albedo do not change depending on the seasons or places. The paper discussed them in Chapter 3.6, but I think this is not a matter of discussion but parts of method (to be discussed in Chapter 2) to proof that estimated values are accurate. Also, the estimation was done by daily or weekly scale. Also, again, please explain how the daily or weekly data are made from original data interval with treatment of missing data (see previous comment!). For instance, daily averaged Tw or Ta is used to obtain Bowen ratio in the formula (4)?
3) I understood that deep lake could store the surface energy till the post monsoon. However, I still wonder that surface evaporation was significantly controlled by the heat storage to determine the lake-level, or not. I thought that lake evaporation would be controlled by surface winds and dryness of atmospheric air mass. Namely, it is a matter of local climate. However, authors suspect it as lake depth matter, such as the importance of heat storage of deep lake (discussing it in Chapter 3.5 based on time lag story). As the Bowen ration is determined by the gradient (4), not by the absolute value of Tw, it is still uncertain that S control post-monsoon evaporation. Wind speed increase after monsoon (Fig.S2) that may be indicating exchanging wet monsoon air-mass to dry one which could mix the lake but also accelerate evaporation. Or, you mean all lakes introduced in L306-313 exist under the same climate and surface-water budget condition? There is still a doubt to L404 “For deep lakes like Paiku Co, contrasting hydrological and thermal intensities determines the large amplitude of seasonal lake-level variations.” It is better to consult with lake hydrologist.

Individual comments

L26　Suddenly started about the “shallow lake”. Why you start implications about deep and shallow matter here? Is this about alpine lake or in general? If there is no clear confidence that lake depth control the lake levels, it is better to delete this part in the abstract.

Abstract; ±** is used. Is this a standard deviation?

L55-65 Better to cite examples of evaporation rate by previous studies here (++mm/day or season) to compare the your result in the abstract.

L97 better to say “daily values were used to calculate*** or estimate +++”.

L114 “radiation at this station is used” for what?

L170-170 There is no information about missing data. Did you conduct perfect observation? Otherwise, better to explain how did you fill or interpolate missing values to get weekly/daily value.

L147 Tw is the same of “surface water temperature” in Fig. 5 ?

L296 “These relationships illustrate how contrasting hydrological and thermal intensities played an important role in the large amplitude of seasonal lake-level variations at Paiku Co.” is not clear.

Hide

RR by Anonymous Referee #3 (01 Mar 2021)

Suggestions for revision or reasons for rejection

The authors revised their manuscript by adding error propagation analysis. Unfortunately, their analysis is not complete. Many of the points which were pointed out in previous reviews were not considered. As a result, I suspect that the magnitude of error is underestimated. Just like the last review, I am in the opinion that this manuscript should be rejected at this point after repeated reviews without successful revision.

1. As pointed out before, the final error estimate should not be made on an annual basis since they are comparing seasonal variations. Errors should be determined each season (month?) based on as much as possible of seasonal (monthly) means/errors.
2. In the equation to estimate the error of Bowen ratio given in L238-239 of their reply, the errors arising from inaccuracy of Ts and es are missing. They should be included. The error of Ts was estimated and is explained in L414-420 of the authors' response. The authors compared 8-day average temperatures. However, since Bowen ratio was calculated daily, the Ts error should also be estimated on daily basis.
3. The error equation is given for the Bowen ratio in the authors' response so that it was easy to check their estimates. However, for other components, methods and equations are not explained. I think the details of error estimates should be provided in their manuscript (perhaps as supplementary material).
3. In L 344- 347 of their main text, they state "The uncertainty in the atmospheric longwave radiation is not estimated, the variations of solar radiation and atmospheric longwave radiation are usually opposite at a site, so their total uncertainty should not exceed the individual uncertainty". I am not familiar with the radiation regime up there, but isn't it unusual that solar radiation and atmospheric longwave radiation change in the opposite direction? When solar radiation increases and air temperature becomes higher with some phase shift, then longwave radiation should also increase. How come it is in opposite direction?
4. The error estimate of the energy storage. As pointed out from the first review, there is another error source: water volume estimate that comes from bathymetry error and water level measurement error. The authors considered the difference between the two lake center temperature profiles. How about likely differences of temperature between at lake centers and near the shorelines?
5. Advection. The authors conclude "Therefore, the impact of river discharge and precipitation on the energy budget at Paiku Co is relatively small and can be neglected." However, the magnitude of river advection is of a similar size to others, and thus the error arising from the omission of the advection term should be included in the error analysis. Note that they did not show evidence to claim the advection due to precipitation can be neglected. They need to show that if they want to say the above statement (this was made in the authors' response in L503-504; this should be mentioned in the main text).
6. Errors in the water balance estimation. "lake evaporation (1.7 mm/day) is similar with the rate of lake-level decrease (1.8 mm/day). " What can be concluded from this? Do not forget about river discharge (0.62-0.89 mm/d) and precipitation (small but not zero in June according to Fig.7). " This difference (0.6 mm/day)....... is very close to the uncertainty of lake evaporation estimated by error propagation. " Do not forget about error of water balance estimation. I think explanations given in the authors' response (L436-447) should also be mentioned in the main text.

Minor comments to the manuscript
L151-152 "There exists surface warming during the day and surface cooling at night for high elevation lakes. However, the daily difference between them is small during most time of a year because surface water can be mixed quickly by water convection or strong wind in the afternoon and the two uncertainties by surface warming and cooling can cancel each other at daily timescale." As pointed out before, these sentences are not appropriate without supporting evidence. In the previous review, I made the same comment. In L330 of the authors' response they write "We agree", and still the same sentences appear in the revised manuscript. I do not understand why.

Minor comments to the responses
L414-420. Not only in the response, but also in the main text, the product's name should be mentioned.
L339-144. "I think it should be 72.8 because it is an integral of water depth." If the equation represents an integral, yes, it should be 72.8. But the equation is not an integral but a summation. Then the number of times you add should be an integer (and not a decimal).
L366-369 "We checked...found the two references are in the reference list"; actually Zhang et al. (2015) was not in the list and is still not in the list. Stainsby et al. (2011) was indeed on the list but in the wrong place (it is still in the wrong place). Livingston (2003) was ok, but it is at a strange place now. I would recommend checking the entire list and correspondence with the main text citation.

Hide

ED: Publish subject to minor revisions (review by editor) (04 Mar 2021) by Bob Su

AR by Yanbin Lei on behalf of the Authors (15 Mar 2021) Author's response Manuscript

ED: Publish subject to revisions (further review by editor and referees) (22 Mar 2021) by Bob Su

AR by Yanbin Lei on behalf of the Authors (26 Mar 2021) Author's response Manuscript

ED: Referee Nomination & Report Request started (11 Apr 2021) by Bob Su

RR by Anonymous Referee #2 (12 Apr 2021)

RR by Anonymous Referee #1 (14 Apr 2021)

ED: Publish subject to minor revisions (review by editor) (24 Apr 2021) by Bob Su

AR by Yanbin Lei on behalf of the Authors (28 Apr 2021) Author's response Manuscript

ED: Publish as is (08 May 2021) by Bob Su

Short summary

Lake evaporation from Paiku Co on the TP is low in spring and summer and high in autumn and early winter. There is a ~ 5-month lag between net radiation and evaporation due to large lake heat storage. High evaporation and low inflow cause significant lake-level decrease in autumn and early winter, while low evaporation and high inflow cause considerable lake-level increase in summer. This study implies that evaporation can affect the different amplitudes of lake-level variations on the TP.