The Fate of Stationary Tools for Environmental Flow Determination in a Context of Climate Change

Round 1

Reviewer 1 Report

Reviewed paper is devoted to the fate of stationary tools for environmental flow determination in a context of climate change. The subject of the paper well fits the scope of Water. The paper could be recommended for publication in Water after revision taking into account the following comments:

1. A comprehensive impact assessment of possible climatic changes is one of the most important problems facing modern science. A large number of different climate models have been developed and analyzed. The reviewed paper provides an overview of studies devoted to a comprehensive analysis of possible changes in flow rates and water temperature in rivers due to climate change. The authors note that the nature of these changes is completely different, when they are reflected in the statistical moments of the first, second, or even more the third order. At the same time, it should be emphasized that the higher the statistical moments are considered, the more complex the procedures for their estimates, the larger the volume of the initial sample is required for their meaningful assessment. The nature of the statistical metrics used is determined primarily by the nature of the tasks solved with their use, therefore, in the analysis of Table. 1 it should indicated for which temperature measurement problems they will be used and which of them are most effective.
2. It should be taken into noted that the considered statistical distribution functions Weibull, Gumbel, Log-Normal, Gamma, Pearson Type III and Log-128 Pearson Type III are at best only some fitting functions that are quite correct in a certain frequency range.
3. The water temperature in the rivers is directly related to the air temperature only during the warm ice-free period. In winter, the ice cover fundamentally disrupts the relationship between air and water temperatures in rivers. Therefore, along with the water temperature, for a more correct analysis, along with the consideration of the links between climatic changes and fluctuations in water discharge, the duration of ice cover period in the rivers should also be considered. This is very important, since the ice cover on the rivers has important ecological functions. Therefore, a combined analysis of changes in water temperature and characteristics of the ice cover seems to be very important.
4. Since the temperature factor is most relevant for rivers located in the permafrost zone, the influence on the formation of runoff, ice phenomena, but most importantly, the intensification of erosion processes is important. Therefore, it is very important and interesting to analyze possible changes in the temperature regime separately for the rivers located in this zone. Moreover, at high latitudes, according to climatic observations and most model calculations, climatic changes are most clearly manifested. These assessments are important not only from the point of view of analyzing the resilience of aquatic ecosystems to possible climate changes, but also for the active economic use of permafrost zones.
5. Despite significant efforts, there is currently no clear understanding of the nature and extent of possible climate change. Under these conditions, the most effective is the analysis of these changes, including the flow rates and water temperature of rivers based on the Bayesian estimates.
6. The concept of nonstationarity of climatic series is very closely related to the time scale of the processes under consideration. Therefore, the recommendation presented in the conclusion that “the metrics should be calculated on the basis of historical data preceding nonstationarity” is incorrect, since the time scale of the phenomena under consideration is not determined.

Author Response

ANSWER TO REVIEWERS

We thank the reviewers for their valuable and constructive comments. Please find our answers below.

REVIEWER 1:

Reviewed paper is devoted to the fate of stationary tools for environmental flow determination in a context of climate change. The subject of the paper well fits the scope of Water. The paper could be recommended for publication in Water after revision taking into account the following comments:

1. A comprehensive impact assessment of possible climatic changes is one of the most important problems facing modern science. A large number of different climate models have been developed and analyzed. The reviewed paper provides an overview of studies devoted to a comprehensive analysis of possible changes in flow rates and water temperature in rivers due to climate change. The authors note that the nature of these changes is completely different, when they are reflected in the statistical moments of the first, second, or even more the third order. At the same time, it should be emphasized that the higher the statistical moments are considered, the more complex the procedures for their estimates, the larger the volume of the initial sample is required for their meaningful assessment. The nature of the statistical metrics used is determined primarily by the nature of the tasks solved with their use, therefore, in the analysis of Table. 1 it should indicated for which temperature measurement problems they will be used and which of them are most effective.

ANSWER:

Table 1 lists minimum eflow metrics that are used by jurisdications, regardless of the thermal regime.  Some text was added to better describe the content of the table :

However, Table 1 provides an incomplete list environmental flow metrics and of jurisdictions that are still using these metrics in hydrological methods for eflows assessment. Although some of these metrics are calculated using seasonal data or monthly data, they can all be potentially used as a minimum flow for the entire hydrological year.

We also agree with the reviewer that higher order moments estimation is challenging. This is now further explained in the paper, below Figure 2:

One potential reason for this exclusion was pointed out by Ulrych et al. (2000), who stated that higher order moments are more difficult to estimate than mean or variance. Their estimation can be more affected by the presence of outliers and small sample size than mean and variance estimators.

1. It should be taken into noted that the considered statistical distribution functions Weibull, Gumbel, Log-Normal, Gamma, Pearson Type III and Log-128 Pearson Type III are at best only some fitting functions that are quite correct in a certain frequency range.

ANSWER:

We agree with the reviewer. We added the following sentences:

Other distributions have been used for low flows. For intance, Shao, Chen, and Zhang (2008) used the Burr III distribution and Liu et al. (2015) used the Generalized Extreme Value distribution. Hence, the selection of the distribution is often contextual and depends on goodness of fit and parsimony, but also on sample size, data quality and the quantiles of interest.

1. The water temperature in the rivers is directly related to the air temperature only during the warm ice-free period. In winter, the ice cover fundamentally disrupts the relationship between air and water temperatures in rivers. Therefore, along with the water temperature, for a more correct analysis, along with the consideration of the links between climatic changes and fluctuations in water discharge, the duration of ice cover period in the rivers should also be considered. This is very important, since the ice cover on the rivers has important ecological functions. Therefore, a combined analysis of changes in water temperature and characteristics of the ice cover seems to be very important.

ANSWER.:

Again, we agree with the reviewer. Text was added to include this information:

Flow is not the only variable that is changing during the winter season. For many rivers, the ice-covered period is shortening, or is predicted to become shorter than at present (Magnuson 2000). Future winter eflows prescriptions will need to account for possible longer ice-free periods in northern rivers.

1. Since the temperature factor is most relevant for rivers located in the permafrost zone, the influence on the formation of runoff, ice phenomena, but most importantly, the intensification of erosion processes is important. Therefore, it is very important and interesting to analyze possible changes in the temperature regime separately for the rivers located in this zone. Moreover, at high latitudes, according to climatic observations and most model calculations, climatic changes are most clearly manifested. These assessments are important not only from the point of view of analyzing the resilience of aquatic ecosystems to possible climate changes, but also for the active economic use of permafrost zones

ANSWER:

Two sentences were added:

This example of Figure 6 is likely representative of thermal patterns for rivers in mid to high latitudes. However, one area for which there is scant information are permafrost rivers. Makarieva et al. (2019) indicated that processes causing changes for basins where permafrost remains dominant are poorly understood. In addition, they stated that small and medium-sized permafrost river basins have been neglected in the literature.

1. Despite significant efforts, there is currently no clear understanding of the nature and extent of possible climate change. Under these conditions, the most effective is the analysis of these changes, including the flow rates and water temperature of rivers based on the Bayesian estimates.

ANSWER :

This is now recognized in the section at the end of section 2, where we added:

Any analysis that is based on climate change scenarios needs to be rooted in a probabilistic framework. Ensemble (i.e. multiple climate scenarios fed to multiple hydrological models) are often used to allow for a probabilistic assessment. Bayesian approaches are often use to implement such frameworks (e.g. Najafi and Moradkhani (2014).

1. The concept of nonstationarity of climatic series is very closely related to the time scale of the processes under consideration. Therefore, the recommendation presented in the conclusion that “the metrics should be calculated on the basis of historical data preceding nonstationarity” is incorrect, since the time scale of the phenomena under consideration is not determined.

ANSWER:

The point made by the reviewer is duly noted. However, we are trying to provide options for managers. The sentence was modified as follows:

The simplest form of adaptation would be to ask that metrics be calculated on the historical data that preceded non-stationarity, provided that this stationary historical time series is sufficiently long to encapsulate the natural variability of the system under study, which can prove to be an important limiting factor.

Reviewer 2 Report

The research topic “environmental flow determination” seems relevance and general interest to the reader of the journal. I found the article to be generally well written, organized, and that gives confident that the authors performed sensible and through field and spectral processing. The designed and structured core objective and achievement of previous works done in the area make paper to be useful. However, following point are needed to be carefully addressed to make the reader friendly before recommended for publication.

Main Points:

1. I have few significant concerns about introduction section of the paper which should be addressed. This is a review article on the recently started research problem (e.g. eflows), However, the author did not include the first definition of the eflows in the introduction. It should be included with concern reference to help audience of the paper in reading and understanding.

2. L88-90. Something does not seem correct. The description in the text does not seem to match with the Figure 2. Please re-write the sentence.

3. L23-236. The point (2) Relative simplicity: What does it mean…? Author did not discuses it prior, more explanation should be added to make it clear.

4. L327- 328. “The sec…, attributes”. I have concerns over this point that what was the key feature were selected group rivers.

5. L378-383. This paragraph of conclusion seems too general. It should be re-written to highlight significance outcomes of the study.

6. Has the author missed the classic work in the field in the references? e.g. (Acreman, Arthington, et al., 2014; Acreman, Overton, et al., 2014; Arthington, et al., 2018)..?

I found many synonyms of the regular words in the manuscript the author should need change regular word instead of synonyms to the words, this might help to make it reader friendly.

Author Response

ANSWER TO REVIEWERS

We thank the reviewers for their valuable and constructive comments. Please find our answers below.

REVIEWER 2:

The research topic “environmental flow determination” seems relevance and general interest to the reader of the journal. I found the article to be generally well written, organized, and that gives confident that the authors performed sensible and through field and spectral processing. The designed and structured core objective and achievement of previous works done in the area make paper to be useful. However, following point are needed to be carefully addressed to make the reader friendly before recommended for publication.

Main Points:

1. I have few significant concerns about introduction section of the paper which should be addressed. This is a review article on the recently started research problem (e.g. eflows), However, the author did not include the first definition of the eflows in the introduction. It should be included with concern reference to help audience of the paper in reading and understanding.

 ANSWER:

Although we understand the point made by the reviewer, we believe that quoting both definitions would be repetitive. Here are both definitions, for comparison:

2007: Environmental flows describe the quantity, timing, and quality of water flows required to sustain freshwater and estuarine ecosystems and the human livelihoods and well-being that depend on these ecosystems.

2018: Environmental flows describe the quantity, timing, and quality of freshwater flows and levels necessary to sustain riverine ecosystems which, in turn, support human cultures, economies, sustainable livelihoods, and well-being.

For the sake of clarity, we prefer only quoting the most recent definition.

1. L88-90. Something does not seem correct. The description in the text does not seem to match with the Figure 2. Please re-write the sentence.

 ANSWER :

We are unsure about what the reviewer is stating here. By our count, the text in the original lines 88-90 is as follows and it is depicted in Figure 2:

However, many authors indicate that the impact of climate change in many regions of the world may be a non-stationarity of order 2: i.e. an increase in variance. In this case, the histogram of daily flows in the future, could be flattened and its breadth increased, compared to the historical time series (Figure 2).

1. L23-236. The point (2) Relative simplicity: What does it mean…? Author did not discuses it prior, more explanation should be added to make it clear.

 ANSWER:

The text was changed to add clarity:

2) Its relative simplicity: all stakeholders easily understand the implications of water warming or cooling. Also, water temperature is often well correlated with air temperature.

1. L327- 328. “The sec…, attributes”. I have concerns over this point that what was the key feature were selected group rivers.

 ANSWER:

The ELOHA framework does not specify which attributes must be selected. It is implied in the methodology that these attributes must allow for relative homogenous groups of rivers.

1. L378-383. This paragraph of conclusion seems too general. It should be re-written to highlight significance outcomes of the study.

 ANSWER:

The conclusion is now more specific, with the aforementioned additional statement and the following additional text:

Although the case study presented in this paper may not be representative of all river ecosystems, it exemplifies how this additional variable may be of great assistance to select an appropriate eflow hydrological metric.

1. Has the author missed the classic work in the field in the references? e.g. (Acreman, Arthington, et al., 2014; Acreman, Overton, et al., 2014; Arthington, et al., 2018)..?

ANSWER:

We thank the reviewer for his suggestions. References were added.

I found many synonyms of the regular words in the manuscript the author should need change regular word instead of synonyms to the words, this might help to make it reader friendly.

ANSWER:

An effort was made to streamline the wording of the manuscript.

Round 2

Reviewer 1 Report

The authors have taken my comments into account in proper way. I recommend this paper for publication.

Reviewer 2 Report

The authors put a lot of energy to revise the manuscript in light of the expert’s comments and suggestions The revised manuscript has been improved. After the careful revision and addressing the reviewer's concerns the revised manuscript is recommended for publication with the following minor changes.

Please revise Figures 1, 2, and 3. Make sure the title of (x and y) axis is visible and the tick marks of the axis should be consistent with Figure 5.