Spatiotemporal Analysis of Evapotranspiration and Effects of Water and Heat on Water Use Efficiency
Round 1
Reviewer 1 Report
see attached file
Comments for author File: 
Comments.pdf
Author Response
Summary of revisions and responses on “Spatiotemporal analysis of evapotranspiration and effects of water and heat on water use efficiency ” (water-1362088)
We appreciate the opportunity to modify our paper according to the critical comments of two reviewers. We have carefully checked and revised the manuscript. Besides, we have polished the language and the charts. This document explains the revisions made in the revised manuscript considering the comments raised by the reviewers. Reviewers’ detailed comments are marked in blue.
The corresponding modifications are indicated by line Numbers in the marked manuscript, and Response messages are marked in black. The revised parts of the manuscript are marked in red.
Reviewers' comments:
Reviewer #1: The aim of the paper is the analysis of evapotransiration and water use efficiency in the arid area of Xinjiang Uyghur Autonomous Region of China.There is no detailed information about land cover and vegetation types. Because a large part of the paper is dedicated to the study of spatial variations, it is needed to provide information about major land-use / land-cover classes and their evolution during the studied period (2000-2014). See for example Yu et al. (2020) who present land-cover in 2000 and in 2018 .
(1) it is needed to provide information about major land-use / land-cover classes and their evolution during the studied period (2000-2014).
Response: Thank you for your comments. We provide information on the major land use/land cover categories and their evolution during the study period (2000-2015).(Figure A1) Due to data limitations, we provide data from 2000 and 2015. We believe that the time difference is one year and the land use type has not changed.(line175-181)
(2) Results about temporal evolution of Ep show a relative stability between 2000 and 2014 (cf. Figure 3). However, Dong et al. (2020) concluded that reference evapotranspiration ET 0 increased at a rate of 3.4 mm/year2 during 1994–2010, a trend explained from the combined effects of increasing air temperature, decreasing relative humidity and net radiation, increasing wind speed. Such a continuous increasing trend in the 21st century has been demonstrated by simulations. As vegetation cover of Xinjiang is generally sparse, the apparent stability of mean Ep during 2000-2014 needs some explanation.
Response: Thank you for your comments. Many thanks to the reviewers for their finalization of this paper and the research results of Dong et al. (2020) mentioned in their opinions. The author argues that there are three differences:
1: The historical observation data of the author in this paper use the National Meteorological Observatory Stations, while the satellite observation data used in this paper have some errors. The value of a raster pixel represents the evapotranspiration of a small area. However, point data can hardly represent the actual situation of small regions (spatial scale difference).
2: Dong et al. (2020) has based on 4 GCMs to forecast the future. the first mode itself exists error, and there is some uncertainty in the simulation region. Secondly, the quantile downscaling method also has some uncertainty. It has not been validated in arid region. Finally, the paper selected RCP4.5. Although it could not represent the real situation of CO2 emissions, some researchers regarded it as closer to the reality.
- The model takes into account various physical processes and many factors. Then, we calculate the observed data based on Hurst method, so there may be some differences between the two results.
Finally, the Hurst method is different from the Mann-Kendall test, and there may be some differences in the results.
In summary, the results confirmed by the simulation need to be further verified by satellite observation and site data, and our results are consistent with those of ADILAI Wufuet al(2017).
(3) Temporal results about the difference between E p and E reflecting drought situation could be compared with the analysis of meteorological drought indices like SPI or PDSI, showing that, on average, drought in Xinjiang has slightly relieved in recent decades, especially in some mountain areas in western and northern Xinjiang (Wang, 2020).
Response: Thank you for your comments. According to reviewers’ comments, we added PDSI and SPEI data, as shown in Figure 3. See the red part of the text for details. (line 120-127 and 251-258)
(4) The comparison show agreements (for example the biggest drought in year 2008) and discrepancies (the wettest year is 2010, but with the difference between E p and E, 2010 appears as a “ normal year”).
Results about WUE response to precicipation and temperature are interesting, with reference to the“warming-wetting”trend in Xinjiang (Wang et al., 2020) : increase in temperature and precipitation since the mid-1980s, with an increase of precipitation in Xinjiang being particularly significant as compared with other regions.
The spatial pattern of pWUE may be compared with the spatial distribution of mean growing season NDVI :
Response: Thank you for your comments. First of all, I am very grateful for the affirmation of this research. Secondly, we have added PDSI and SPEI trend graphs. From Figure 3, we can see that the driest in 2008 and the wettest in 2003. This result is verified in Figure 3c and d. The results are basically the same. We analyzed the spatial distribution of NDVI in the growing season and pWUE and the reasons for the differences. . See the red part of the text for details. (line 367-381)
(5) It is recommended to present WUE annual or seasonal maps with a detailed color scale between 0 and 1, masking the higher values.
Response: Thank you for your comments. On the basis of the reviewers’ comments, We modified the Mean annual of WUEs in Figure 6. We used the normalization method to map the values between 0 and 1, and the color scale of the map was between 0 and 1, which masked the higher values. We have also revised the corresponding text in the manuscript.See the red part of the text for details. (line 359-366)
(6) The differences observed in the responses of WUE to precipitation between northern and southern Xinjiang could be quantified and further discussed with reference to land cover changes in the last decades
Response: Thank you for your comments. We added the land use cover data from 2000 to 2015, combined with the land use information, analyzed the regions and distribution with large ET fluctuations, and then analyzed the causes.
We are deeply sorry for our experts' opinions on quantifying and further discussing the difference in response of water use efficiency to precipitation between southern xinjiang and northern Xinjiang based on land cover changes in the past decade, mainly for three reasons:
- The topic of our paper is the temporal and spatial variation of water use efficiency of three different hydrological indicators. The analysis of the difference between northern xinjiang and southern Xinjiang based on land cover change has no contribution to the topic of the paper.
- The WUE spatial resolution calculated by us is 25km, and most of southern Xinjiang is null value area, so the statistical results are too uncertain.
- According to the research results (Figure. 9d, e, f), precipitation has a negative effect on WUE. If it is quantified separately in southern xinjiang and northern Xinjiang, we think there is not much difference.
(7) Typing errors : sTableility (line 76). sTableility (line 163) sTablele (line 167) unsTablele (line 230) Stableility (lines 235 and 236) sTablele, and the sTableility (line 253)
Response: Thank you for your comments. According to the Reviewers' comments. We modified all typing errors in the Manuscript, See the red part of the text for details.
.(In the reference section, blue marks are added references, red marks are revised documents)
Author Response File: Author Response.pdf
Reviewer 2 Report
Review WATER 1362088 for peer review
Title: Spatiotemporal analysis of evapotranspiration and effects of water and heat on water use efficiency
Yuan yuan Tang et al
General comment
The paper presents a spatial and temporal evaluation of water use efficiency of the arid to semiarid large Xinjiang region in NW China. This topic is relevant, and for sure to this water scare region. Several (three WUE) indices are defined and submitted to an analysis. Also evapotranspiration derived from GLEAM v3. Global dataset is analyzed and used. The authors made a good attempt to publish their research, but a more clear description of methods and interpretation results is needed.
Although lot’s a data analysis is done, several parts of the analysis and definitions are poorly described, I have some doubts with some outcomes and figures and lots of language errors are found.
I go here through the paper.
Specific comments
The abstract contains plenty of writing errors. E.g. line 19 “different hydrological processes and its response to Hydrologicalprocesses” ???...
Introduction: In the introduction, a good try is done to define and describe the problem and objectives. Also annoying language errors “Line 76: “sTableility” are visible.
Materials and methods:
This starts good.
Figure 1: Legend is probably of all the P.R.China with 8500 m max. Please stretch this picture and show correct legend (reduce the range for Xinjiang only).
Data are reasonably described (sources, etc.).
On Page 4, Figure 2, the authors claim a good relationship (regression); but I see in fct 2 populations of data, and a number (> 10) lower outliers. Is there an explanation for these observations?
2.3 PCP and TMP data and also MODIS GPP an GLEAM
I’m surprised to see that the authors have e.g. many 1-km and other good spatial resolution data. But they resample all to 0.25 degrees (~25 km). This is of course, due to the GLEAM resolution probably; but a pity.
2.5 CV
The CV definition is known by any scientist. I don’t see the need to define it with an equation etc.? Line 159-160.
Lines 163, 167: language errors………
2.6 Hurst index: 0k
2.7 Calculation of WUEs
They should be a bit more careful here, in presenting the three WUEs. Nowhere units and also ET stands for Precipitation etc.. ET stands for GLEAM-Et etc. Be more specific.
Also the spatial (0.25 Degree) and time scale (interval) of the data (monthly? ) and WUE indices…
Equation 8: linear trend – slope…. This is not needed. Most people know a linear regression concept.
2.8 WUE response (pls. also error in title)
“We used Rstudio tool to detrend all data”… (on line 201..). This is important how this was done.
It is not Rstudio, but probably a R package. Pls. mention it + method applied.
Line 203> How are two different variables Precipitation and temperature merged in 1 variable (see equation 9). ???
Line 238: error STableility…..
And more in text.
- Results and discussion
In 3.4 on the discussion of the WUE results, finally, some mention on the values is given (line 296).
Because three WUEs are used with different denominators (ET, Gleam-T and PCP) is is very difficult for the reader to follow. This part is interesting, but not so well described.
3.5 WUE and PCT/TEMP thresholds
Figure 8 is largely misleading I find. GPP has thresholds for both temperature and precipitation (moisture supply). Also a WUE must be evaluated within reasonable thresholds. Otherwise, the concept fails (or is misapplied).
Example, at freezing temperature, we may have ice/snow sublimation, but little (zero) evapotranspiration or plant transpiration. So, you WUE=GPP/ET will get relatively high, because the denominator -> zero.
We all know that GPP (gross primary productivity and also plant transpiration become very very low and almost zero at cold (e.g. below freezing temperatures).
If you then compute a WUE (with a very low denominator), you will get a positive value, even if the GPP is almost zero.. A bit a strange way
3.6 Spatializing
This is an interesting way to show local/regional differences of the WUEs etc.
Conclusion
The research idea behind this paper is good, but in total, the data analysis, but surely how it is all presented, interpreted and discussed is not yet ok.
I therefore recommend a major revision.
Authors should revisit the WUEs presented, and give more quantitative explanation on their behavior (vs. PCP and TMP), also specify units (because ET is replaced at hoc by precipitation etc….). This remains confusion to the reader.
Final comment.
I recommend a major revision for the authors to react to the suggestions made.
Author Response
Summary of revisions and responses on “Spatiotemporal analysis of evapotranspiration and effects of water and heat on water use efficiency ” (water-1362088)
We appreciate the opportunity to modify our paper according to the critical comments of two reviewers. We have carefully checked and revised the manuscript. Besides, we have polished the language and the charts. This document explains the revisions made in the revised manuscript considering the comments raised by the reviewers. Reviewers’ detailed comments are marked in blue.
The corresponding modifications are indicated by line Numbers in the marked manuscript, and Response messages are marked in black. The revised parts of the manuscript are marked in red.
Reviewers' comments:
Reviewer #2: The paper presents a spatial and temporal evaluation of water use efficiency of the arid to semiarid large Xinjiang region in NW China. This topic is relevant, and for sure to this water scare region. Several (three WUE) indices are defined and submitted to an analysis. Also evapotranspiration derived from GLEAM v3. Global dataset is analyzed and used. The authors made a good attempt to publish their research, but a more clear description of methods and interpretation results is needed.Although lot’s a data analysis is done, several parts of the analysis and definitions are poorly described, I have some doubts with some outcomes and figures and lots of language errors are found.I go here through the paper.
(1) Specific comments:The abstract contains plenty of writing errors. E.g. line 19 “different hydrological processes and its response to Hydrological processes” ???...Introduction: In the introduction, a good try is done to define and describe the problem and objectives. Also annoying language errors “Line 76: “sTableility” are visible.
Response: Thank you for your comments . We have changed “in order to reveal the variation rule of WUEs different hydrological processes and its response to Hydrological processes” to “in order to reveal the variation rule of WUEs based on hydrological indicators and its response to climate” (Lines19-20). I sincerely apologize for the language errors in the manuscript. We have corrected similar language problems in the full text. Besides, we have polished the language.
(2) Materials and methods:
Figure 1: Legend is probably of all the P.R.China with 8500 m max. Please stretch this picture and show correct legend (reduce the range for Xinjiang only).
Data are reasonably described (sources, etc.).
On Page 4, Figure 2, the authors claim a good relationship (regression); but I see in fct 2 populations of data, and a number (> 10) lower outliers. Is there an explanation for these observations?
Response: Thank you for your comments. We have modified Figure 1 and the color in the figure ,reduce the range for Xinjiang. We have revised the name of the China meteorological Data Network(Line 104-105,Line130).
For the validation data part, as the scatter plot shows that there are some outliers. For this phenomenon, we believe that we did not delete the site data when verifying GLEAM data. If we delete the site data, we think it will increase the human factor, resulting in higher accuracy of verification, which is not what we want. We want to show the true situation. It can be seen from Figure 1 that some of the sites we selected are located in deserts and some are located in mountains, and Xinjiang has a large area, which must lead to the existence of outliers in some sites. For verification results, we mainly look at the overall effect, not the site or specific region. Therefore, we believe GLEAM data can be used in Xinjiang.
(3) 2.3 PCP and TMP data and also MODIS GPP an GLEAM ,I’m surprised to see that the authors have e.g. many 1-km and other good spatial resolution data. But they resample all to 0.25 degrees (~25 km). This is of course, due to the GLEAM resolution probably; but a pity.
Response: Thank you for your comments. First of all, thank you very much for your affirmation, and we are deeply sorry that we have to do so in order to match GLEAM data resolution. Secondly, this is only an attempt mainly on spatial variation and response. In the future, we will try to carry out high resolution simulation based on CLM to carry out similar work.
(4) 2.5 CV:The CV definition is known by any scientist. I don’t see the need to define it with an equation etc.? Line 159-160.Lines 163, 167: language errors………
2.6 Hurst index: 0k
Response: Thank you for your comments. We've removed equation (1). We believe that Hurst index method is more complicated than CV method, and it is difficult for ordinary readers, so we keep the equation of Hurst index method, so as to improve the readability of the article. I sincerely apologize for the language errors in the manuscript. We have corrected similar language problems in the full text. Besides, we have polished the language.
(5) 2.7 Calculation of WUEs
They should be a bit more careful here, in presenting the three WUEs. Nowhere units and also ET stands for Precipitation etc.. ET stands for GLEAM-Et etc. Be more specific.Also the spatial (0.25 Degree) and time scale (interval) of the data (monthly? ) and WUE indices…
Equation 8: linear trend – slope…. This is not needed. Most people know a linear regression concept.
Response: Thank you for your comments. First, we modified equation (7), and then detailed the data used to calculate WUE (including: GPP, precipitation, spatial resolution and temporal resolution)(lines219-221). Finally, we removed equation (8).
(6) 2.8 WUE response (pls. also error in title)
“We used Rstudio tool to detrend all data”… (on line 201..). This is important how this was done. It is not Rstudio, but probably a R package. Pls. mention it + method applied. Line 203> How are two different variables Precipitation and temperature merged in 1 variable (see equation 9). ??? Line 238: error ST ableility…..And more in text.
Response: Thank you for your comments. According to the opinions of reviewers, we have modified line201, detailing the calculation process and methods used. Secondly, for precipitation and temperature, the expression of the article is not clear due to the spelling errors of the author. We apologize for the inconvenience brought to you , we have modified (lines229-238). We modify all Typing errors in the Manuscript; See the red part of the text for details.
(7) Results and discussion
In 3.4 on the discussion of the WUE results, finally, some mention on the values is given (line 296).
Because three WUEs are used with different denominators (ET, Gleam-T and PCP) is is very difficult for the reader to follow. This part is interesting, but not so well described.
Response: Thank you for your comments. First of all, thank you very much for the reviewer's affirmation. We have combed and revised the whole text of this part. For details, please see the red part of the manuscript.
(8) 3.5 WUE and PCT/TEMP thresholds
Figure 8 is largely misleading I find. GPP has thresholds for both temperature and precipitation (moisture supply). Also a WUE must be evaluated within reasonable thresholds. Otherwise, the concept fails (or is misapplied).
Example, at freezing temperature, we may have ice/snow sublimation, but little (zero) evapotranspiration or plant transpiration. So, you WUE=GPP/ET will get relatively high, because the denominator -> zero.
We all know that GPP (gross primary productivity and also plant transpiration become very very low and almost zero at cold (e.g. below freezing temperatures).
If you then compute a WUE (with a very low denominator), you will get a positive value, even if the GPP is almost zero.. A bit a strange way
Response: Thank you for your comments. First of all, thank you very much for the expert's discussion on this issue. For example, at freezing temperature, snow sublimation, evapotranspiration and transpiration are very few, leading to high WUE, or GPP is very low at temperature, almost zero. First of all, there have been similar studies on WUE in Xinjiang, whose findings are almost consistent with our results[1]. Secondly, when we screened the data, we found that evapotranspiration in some desert areas was null. Finally, we used the monthly mean value to calculate, ignoring the extreme phenomenon you mentioned. If we follow the advice of the experts, we should divide regions, such as frozen regions, wet regions and so on, which further narrows the area we study. The spatial resolution of our data cannot meet the requirements, even if it is divided into regions, it will cause great uncertainty.
(9) 3.6 Spatializing
This is an interesting way to show local/regional differences of the WUEs etc.
Response: Thank you for your comments. Thank you very much for the expert's affirmation, Besides, we have polished the language.
(10) Conclusion
The research idea behind this paper is good, but in total, the data analysis, but surely how it is all presented, interpreted and discussed is not yet ok.
I therefore recommend a major revision.
Authors should revisit the WUEs presented, and give more quantitative explanation on their behavior (vs. PCP and TMP), also specify units (because ET is replaced at hoc by precipitation etc….).
This remains confusion to the reader.
Response: Thank you for your comments. Thank you very much for the experts' affirmation. We made a detailed explanation of the data analysis part and discussed the results. In addition, we specify the units of the WUE that we define, and finally we polish the full language.
Response: Thank you for your comments. The author supplements the actual references with the reviewer's comments and author's guide, and the author reviews the complete bibliography.
.(In the reference section, blue marks are added references, red marks are revised documents)
References
[1]ADILAI Wufu, YUSUFUJIANG Rusuli et al Spatio-temporal distribution and evolution trend of evapotranspiration in Xinjiang based on MOD16 data . Geographical Research,2017,36(7),1245-1256.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The manuscript has been greatly improved through additions of new data and comments. I recommend publication in present state.
