Increased surface water evaporation loss induced by reservoir development on the Loess Plateau

Liu, Yao; Xie, Xianhong; Wang, Yibing; Tursun, Arken; Peng, Dawei; Wu, Xinran

doi:https://doi.org/10.5194/egusphere-2025-11

Preprints

https://doi.org/10.5194/egusphere-2025-11

Preprints

12 Mar 2025

| 12 Mar 2025

Increased surface water evaporation loss induced by reservoir development on the Loess Plateau

Yao Liu, Xianhong Xie, Yibing Wang, Arken Tursun, Dawei Peng, and Xinran Wu

Abstract. Global-scale reservoir construction has significantly enhanced local water supply for local production and livelihoods, yet the evaporation losses from these surface water bodies remain poorly understood, particularly in the context of climate change. The majority of existing studies have predominantly focused on terrestrial evaporation, overlooking the intricate evaporation dynamics within these aquatic systems. This study addresses this gap by investigating water body evaporation in the Loess Plateau of China, a region characterized by extensive reservoir development over the past decades. By employing a modified Penman equation and utilizing long-term remote sensing water body data to calculate water depths while accounting for the thermal storage capacity of water bodies, we estimated water evaporation rates and total evaporation volumes for the period 2000–2018. Validation against pan evaporation observations demonstrates the efficacy of our improved approach in capturing the evaporation patterns of diverse water bodies in the Loess Plateau. Results reveal a subtle decreasing trend in evaporation rates across the region. However, the total evaporation volume amounts to a substantial 4.16×10⁶ m³/d, with a notable upward trend at a rate of 0.117×10⁶ m³/d/yr. Attribution analysis shows that while the combined effects of climate change marginally reduced evaporation rates, the expansion of water bodies has counteracted this trend, resulting in a significant increase in total evaporation losses. Particularly, the development of small- and medium-sized reservoirs and check dams is the primary driver of increased evaporation losses on the Loess Plateau. Given comparable evaporation losses to surface water withdrawals in this region, future water management and hydraulic projects must consider such substantial losses. This study fills gaps in evaporation dynamics and underscores the need for integrated strategies addressing climate change, reservoir expansion, and evaporation.

Received: 02 Jan 2025 – Discussion started: 12 Mar 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Yao Liu, Xianhong Xie, Yibing Wang, Arken Tursun, Dawei Peng, and Xinran Wu

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-11', Anonymous Referee #1, 14 Apr 2025
The manuscript titled ‘Increased surface water evaporation loss induced by reservoir development on the Loess Plateau’ by Yao Liu et al., presents an analysis of evaporation rates and total evaporation volumes from surface water bodies across the Loess Plateau. The study investigates the temporal evolution of these components over the past two decades, explores their spatial distribution, and performs attribution analysis to identify the key drivers of changes in open water evaporation volume.
The objectives of the study are well described, and the results are thoroughly discussed. A key finding of this study is the paradoxical behavior of decreasing evaporation rates, yet with increasing evaporation volume on the Loess Plateau, whereby mainly the increase in area in small- to medium-sized water bodies over the last decades contributed to this increase in evaporation volume.
Below, I provide several comments and suggestions that may help improve the impact of the manuscript:
Be cautious when drawing conclusions regarding increases in total evaporation losses over the region, due to the expansion of water body area. Take into account that with an increased extent of water body area over time, the open water evaporation amount will indeed rise. This relationship should be clearly acknowledged when interpreting the results.

In the introduction it is highlighted that terrestrial evaporation increased over the past years, where some key climatic factors play a crucial role. One would expect that these variables also induce rises in open water evaporation rates over the area, however, the results of this study indicate a downward trend in evaporation rate. Could the authors further elaborate on the mechanisms behind this trend?

The authors state to have developed an improved approach for estimating open water evaporation, which accounts for the thermal storage capacity of water bodies. However, no comparison or evaluation against other existing methods is presented. Consequently, the conclusion made on page 41, lines 13–15, should be reconsidered or more carefully justified.

Related to the point above: on page 8, line 1, the introduction refers to an "enhanced methodology." Please clarify in what way the approach is enhanced, especially since the described methodology has already been applied in similar studies.

Data and Methods: Please explicitly state the study period, as well as the temporal and spatial resolution of the datasets and evaporation estimates in this study.

The algorithm for water depth estimation is not validated in this study. However, based on Eq 6, the water depth directly affects the calculation of thermal storage and, by extension, evaporation rates.

Additionally, the assumption that the slope of the water body is equivalent to that of its surroundings is quite strong. Was any sensitivity or uncertainty analysis conducted to assess how this assumption might influence heat storage and subsequent evaporation estimates? Please elaborate.

Regarding the validation of evaporation trend and volume for the individual pans: in what way are the chosen pan coefficients influencing the validation results? Have other coefficients been tested?

Is there a specific reason for using varying significance levels in different analyses (e.g., in Figure 8)? To facilitate comparability, it is recommended to apply a consistent significance level across all variables and figures.

Choice of colormaps: Please ensure that appropriate colormaps are used for different types of data. Diverging colormaps are suitable for showing changes, anomalies, or trends, but sequential colormaps are more appropriate for depicting absolute values, such as evaporation rates or volumes. Updating the colormaps accordingly would improve visual clarity.

Throughout the manuscript biases are reported. Be cautious when reporting average or median biases if the dataset includes both positive and negative values. Instead, it is more informative to report the range or the mean absolute bias.

Attribution analysis: results illustrate that water body area is the main driving factor for increased evaporation volumes over time, while the contribution of climatic drivers was found to be very small. Given the magnitude of changes in water body area, it is expected that this variable will take over the influence of the other climatic drivers. It would be interesting to perform a similar attribution analysis on evaporation trends (excluding the area) to identify the main driving factor for the general decrease in open water evaporation rates over the region. Additionally, a spatial representation of the contributions of the different forcing variables would provide valuable insights. It would be interesting to show the variable with the highest contribution per pixel. This way, spatial patterns in contribution can be assessed.

A concern arises regarding the attribution analysis: SSR was found to be the main contributing climatological factor to the increasing trend in evaporation volumes. However, a downward trend in SSR was observed (Fig 8f), which corresponds to a decrease in evaporation rates. Could the authors clarify this contradiction, and explain how the results of the contribution analysis should be interpreted?

Discussion: page 39 line 16: it is mentioned that the JRC-GSW data show data gaps. Please clarify in Section 2.2 how these gaps were handled. If gap-filling was applied, the method should be described.

The authors emphasize the utility of their findings for regional or local water management. Please be more specific about how the results can be applied in practice. Consider rephrasing relevant parts of the conclusion (e.g., page 41, lines 21–22; page 42, lines 1–2) to reflect this more concretely.

Please improve the writing and refine the language of the manuscript, in order to improve the overall clarity and readability.

In addition to this, I have identified several minor comments:
Abstract: page 3, lines 1-2: ‘This study fills gaps in evaporation dynamics’. How exactly is this study filling gaps? Please rephrase.

Introduction: page 5 lines 20-22: Please rephrase and explain better what those projects are.

Introduction: page 4 lines 18-19: Wind speed is not mentioned here as a driving factor, yet further in the study wind speed is considered as important.

Introduction: page 5 line 21: Remove ‘in’.

Introduction: page 6 second section: Pan and eddy-covariance measurements are a type of in-situ observations. On the other hand, evaporation estimates are made by models, which can be validated using observations. Please consider rephrasing lines 9-14 to make the distinction clear.

Introduction: page 7 lines 9-11: Please rephrase the sentence.

Introduction: page 8 line 5: Remove the word ‘primary’ as you don’t have any secondary objectives.

Introduction page 8 line 8: Consider removing the second part of the sentence of the 2^nd objective, as it already illustrates the outcome of the experiment.

Data and Methods: page 9 line 11: Use ‘e.g.’ instead of ‘etc.’

Data and Methods: Fig 1: Please specify what the fraction water body is, is this water body area per pixel?

Data and Methods: page 10 line 5: The reference is from the year 2016, although the assessment goes until 2018. Please provide a correct reference, and mention the studied period.

Data and Methods: page 1 line 13: What does ‘reliable’ mean in this sentence?

Data and Methods: page 10 lines 19 and 22: Please provide correct citations, no urls in the manuscript text.

Data and Methods: page 10: Have all data been converted to the same spatial resolution? What is the final spatial resolution used for evaporation estimates?

Data and Methods: page 11 line 1: For completeness, please provide the diameter of the big pans.

Data and Methods: page 11 line 2: Please specify which type of difference? Is it the size or location?

Data and Methods: page 12 line 10: How is the prevailing wind direction calculated?

Data and methods: page 12 Eq (3): κ↓and α are not described.

Data and Methods: page 15 Fig 2: The figure caption is not clear. Please rephrase.

Data and Methods: page 16: Please provide references for attribution analysis equations.

Results: Section 3.1: Is there a particular reason for reporting the validation results in monthly resolution, while for the other results, evaporation rates and volumes are reported with a daily resolution?

Results: Fig 3 and 4: Please consider adapting the layout of the plots, making them bigger in the horizontal direction, and consider reporting correlation, bias, and RMSE in individual plots.

Results: Section 3.1: When showing time series of individual pans, please add a short description of the performance of the individual pans in the text.

Results: Fig 6c and 7d: What do the shaded areas represent? Please add to the figure caption.

Results: caption of Figures 6 and 7: Please add more detail to figure caption: ‘long-term average’: how long is this exactly, which years are taken into account for calculating the trend and the climatology?

Results: Fig 7c: Note that compared to Fig 6c where an average E rate for the Loess Plateau was plotted, here the total sum of daily evaporation volume over the entire Loess plateau is shown, please specify in the figure caption.

Results Fig 8: Spatial patterns: Please specify the period for the trend analysis in the caption.

Results: Fig 8: Temporal changes: Please specify units for trend (consistency with other plots in manuscript). When significant, please indicate the significance level (e.g. Fig 8b). Note the different significance levels in this plot, aligning with a general comment above. Describe the meaning of the colored band in the caption.

Results: Fig 9a: Use a sequential colormap instead. Is the reported water area, the water area per grid cell? Please specify in the caption.

Results: Fig 9: In the caption it is mentioned that the grid is 0.05°. This does not correspond with the grid mentioned in the Methods part (30m). Please clarify the grid size difference.

Results: Fig 9: Bar plots: Consider using percentages instead of counts, as percentages are also reported in the text.

Discussion: page 33 lines 19-22. The comparison with the study of Zhang et al.: What is the period taken into account in their study? This period might be important in framing conclusions given the strong trend found in this study. Is the area in the study of Zhang et al. the same as the area studied here? As you’re looking at total evaporation volumes per day for a certain area, this is important to take into account.

Discussion: page 38 Fig 12: Please provide a reference for the water withdrawal data, and report in Section 2.2.

Discussion: page 38 Fig 12: The authors could indicate the 100% line, indicating the level where evaporation volumes equal the water withdrawal. When this line is crossed more water evaporates than is withdrawn, which is happening the last 5 years, and is of major importance for water policy in this region.

Conclusion: page 41 line 4: The peak in evaporation trends did not occur in May and October, but instead in July and August. It is evaporation volume that shows peaks in May and October.

Conclusion: page 41 lines 19-20: ‘This study thus demonstrates that the research methods employed are readily extendable to other regions.’ Note that this is not demonstrated in this study.
Citation: https://doi.org/10.5194/egusphere-2025-11-RC1
- AC1: 'Reply on RC1', Xianhong Xie, 11 May 2025
  
  We sincerely appreciate the valuable feedback from the anonymous reviewer on our manuscript. Our detailed response to the reviewer's comments, including the planned modifications to the manuscript, is attached in the attached file.
  
  Citation: https://doi.org/10.5194/egusphere-2025-11-AC1
RC2:
'Comment on egusphere-2025-11', Anonymous Referee #2, 24 Apr 2025
Liu et al. improved the Penman equation by considering water depth to calculate evaporation rate and volume for the Loess Plateau. They also performed an attribution analysis using climate, radiation, and area factors. They found that while evaporation rates decreased, the evaporation volume showed an increasing trend. Their analysis suggested that the increase in water body area offset the decrease in evaporation rates, leading to an overall increase in regional evaporation. This work provides an interesting and novel perspective on the causes of changes in regional evaporation. However, the manuscript still requires some revisions, as detailed below:
Major Comments:
In arid and semi-arid regions, such as the Loess Plateau, a large number of small water bodies change relatively quickly. Can the lag time in equations 4 and 5 adequately account for the impact of the variation of small water bodies?

The resolution used for the attribution analysis is coarser than the previous spatiotemporal results. How was the resolution conversion handled?

How much improvement in accuracy does the modified method provide compared to the Penman equation that does not consider water depth?

It would be beneficial to include comparisons with results from other regions in the discussion. For instance, the introduction mentions examples from the U.S. and the Tibetan Plateau, where changes in water body area due to climate change differ from the human-induced expansion of water bodies on the Loess Plateau.

The writing could be more concise. There are unnecessary explanations in the results, such as on pages 19 (lines 4-12, which should be moved to the discussion or removed), 25 (lines 3-7, which should be moved to the discussion), 22 (lines 9-11, recommended for removal), 23 (lines 2-4, recommended for removal), and 26 (lines 2-5, which can be shortened).

In the second paragraph of the introduction, the examples from the U.S. and the Tibetan Plateau focus on climate-driven changes in water area, which ultimately cause changes in ET. It would be more appropriate to use examples of human-induced changes in water area.

The section describing the study area should include an overview of ET in this area.

Minor Comments:
In Fig. 1a, the legend for "River" should be a line; in Fig. 1b, the text is too small. Also, there are capitalization errors in the figure title.

Validation metrics such as R², RMSE, and Bias should be briefly explained in the methods section, after the evaporation pan data description, without listing the formulas.

Page 19, line 2. The full forms of R² and RMSE have already been provided in lines 6 of page 18, so there is no need to repeat them.

In Figs. 3 and 4, it is recommended to label R², RMSE, and Bias on the figures to provide readers with overall validation information.

In Fig. 5, using the same color for "Pan" and "Big pan" would improve readability. They can be distinguished by different symbols, and it may be better to use darker colors to represent better results.

In Fig. 6 and 8, it is recommended to use pie charts to show the proportion of area increases and decreases, or frequency distribution charts for each classification.

How are large, medium, and small water bodies defined?
Citation: https://doi.org/10.5194/egusphere-2025-11-RC2
- AC2: 'Reply on RC2', Xianhong Xie, 19 May 2025
  
  We are grateful to the distinguished anonymous reviewer for the thoughtful and constructive feedback provided on our manuscript. Our detailed response to the reviewer's comments, including the planned modifications to the manuscript, is attached in the attached file.
  
  Citation: https://doi.org/10.5194/egusphere-2025-11-AC2

Yao Liu, Xianhong Xie, Yibing Wang, Arken Tursun, Dawei Peng, and Xinran Wu

Data sets

Land surface water evaporation during 2000-2018 on the Loess Plateau Yao Liu https://doi.org/10.5281/zenodo.14963639

Yao Liu, Xianhong Xie, Yibing Wang, Arken Tursun, Dawei Peng, and Xinran Wu

Viewed

Total article views: 969 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
814	133	22	969	20	34

HTML: 814
PDF: 133
XML: 22
Total: 969
BibTeX: 20
EndNote: 34

Views and downloads (calculated since 12 Mar 2025)

Month	HTML	PDF	XML	Total
Mar 2025	57	16	2	75
Apr 2025	62	42	6	110
May 2025	44	14	4	62
Jun 2025	53	6	3	62
Jul 2025	41	12	0	53
Aug 2025	126	13	5	144
Sep 2025	403	20	1	424
Oct 2025	28	10	1	39

Cumulative views and downloads (calculated since 12 Mar 2025)

Month	HTML	PDF	XML	Total
Mar 2025	57	16	2	75
Apr 2025	62	42	6	110
May 2025	44	14	4	62
Jun 2025	53	6	3	62
Jul 2025	41	12	0	53
Aug 2025	126	13	5	144
Sep 2025	403	20	1	424
Oct 2025	28	10	1	39

Viewed (geographical distribution)

Total article views: 992 (including HTML, PDF, and XML) Thereof 992 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 23 Oct 2025

Short summary

Global reservoir construction has improved water supply, but its impact on evaporation losses remains uncertain. Our study examines how reservoir development affects water evaporation in the Loess Plateau of China. We find that total evaporation has increased due to expanding water surfaces, particularly from thousands of small- and medium-sized reservoirs. This new finding underscores the need to consider previously overlooked evaporation in the hydrological cycle and future water management.


Total:	0
HTML:	0
PDF:	0
XML:	0