the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Oct 2024
Constantly renewing glacial lakes in the Kyrgyz Range, northern Tien Shan
Abstract. Glacial lakes in the Kyrgyz Range of the northern Tien Shan, Kyrgyz Republic, Central Asia are monitored due to concern over possible glacial lake outburst floods (GLOFs) after recent glacier shrinkage in the area. To evaluate the status of these lakes, we investigated the number of glacial lakes and the lake area from 1968 to 2021 using Corona KH-4, Landsat (7 to 10), Sentinel-2, and PlanetScope satellite images. We found that the number of glacial lakes increased by 30 % from 417 in 1968 to 543 after 2000. However, some lakes vanished. In particular, 305 of the original 417 glacial lakes vanished by 2000, whereas 431 of the 543 glacial lakes identified between 2000 and 2021 formed after 2000. In addition, the total lake area significantly increased from 0.87 km2 in 1968 to 5.21 km2 in 2021. New glacial lakes rapidly formed because the glacier area has decreased by 32 % over the past 50 years, and the places from which glaciers retreated have changed into glacier-moraine complexes (GMCs) on which new lake basins appeared and expanded. Thus, the renewal of glacial lakes in the Kyrgyz Range is the outcome of glacier shrinkage and ice melting within GMCs.
- Preprint
(1396 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2024-160', Anonymous Referee #1, 24 Nov 2024
This study does two main studies: the first study is the history of glacial lakes and changes in recent years; the second study is the changes in the glacial-moraine complex. The paper specifically has some scientific significance, but there are some serious problems that need to be improved.
- In response to the introduction of the data sources for the first study, it is stated at the end of the introduction what data we have used for this study. And in the Methods section, only some information about the Corona KH-4 data is presented, not the other image data, such as the specific time of the image, the quality status of the image, the spatial resolution of the image, the pre-processing of the image, and the source of the data. For studies that use multiple sources of data, it is best to use a tabular format to present a summary of the underlying information for each data source.
- In section 3.2, it was mentioned that glaciers and glacial lakes from all data sources are extracted by manual outlining, how can the glacial lake boundary errors be taken into account? Due to the large difference in spatial resolution of different image sources, how to resolve the relationship between the boundary accuracy extracted from different resolution images and the actual changes of the glacial lake. It is mentioned here that a threshold of 100 m2 is used to extract a glacial lake, which is just one image pixel for a Sentinel-2 image, how can we be sure that this is a glacial lake for such a small image pixel? What is the definition of an glacial lake for this study? With such a small area threshold, the spatial resolution of the image in Landsat 7/8 imagery, even after panchromatic band fusion, is only 15 m. With an area of 225 m2 for one image pixel, how is that very small glacial lake recognizable in Landsat 7/8 imagery? In addition, no information is provided in this section on what bands of imagery were used to manually outline the glacial lake and what kind of pre-processing was applied to the imagery.
- In section 3.3, detailed information on the specific methodology of image processing for analyzing changes in GMCs in the study area using 49-view ALOS-2/PALSAR/-2 imagery is indicated in a number of previous studies. This write-up is too simple and broad and the key processing steps should be described in the methods section of this paper. Overall, the authors' presentation of the data and methodology is too brief and not very descriptive of the study. The author used DEM data from four different sources to study changes in the lake basin surface. Are the elevation datums consistent across these four data sources? What are their respective spatial and vertical resolutions? How did the author address the issue of matching between the different DEM data sources? These data processing details are essential; otherwise, the analysis results may be unconvincing.
- In Section 4.1, it is mentioned that there are 543 glacial lakes during the period 2000–2021. This is difficult for me to understand, as it spans a 22-year time frame. How was it determined that there are 543 glacial lakes over such a long period? Glacial lakes tend to change rapidly, so it is more common to describe the number of glacial lakes at a specific year rather than over such an extended period like 2000–2021.
- In the text, many superscripts for "km2" are not written correctly.
- What are the identification criteria for GMCs? Please clearly specify in the text how the 611 GMCs were determined.
- I cannot understand 'whereas the areas of lake basins in 2021 increased eightfold compared with 1968.' Please provide an explanation of this sentence
- In line 144, “Along a cross-section, a comparison of DEM differences (Fig. 6a) between 1991 and 2013 show…”. It should be “shows” not “show”.
- Line 163-164: “Consistent with the glacier trends there, the glacial area here decreased by 30% over the past 53 years.” In this sentence, what is the difference between “glacial area” and “glacier”?
- In section 5.1, a variety of scenarios of rapid glacial lake formation are listed, please support these scenarios by showing them in the form of remote sensing image maps in the context of the study area.
- In the conclusion, the author identifies 611 GMCs, most of which contain dead ice. How was dead ice identified within these GMCs?
- In the second half of the conclusion, the quantification of the findings is insufficient.
Citation: https://doi.org/10.5194/nhess-2024-160-RC1 -
RC2: 'Comment on nhess-2024-160', Jan-Christoph Otto, 25 Nov 2024
The authors present a study on glacier lake evolution in the northern Tien Shan mountains based on multiple period remote sensing data. They applied manual mapping methods and used DEM-of-difference analysis to quantify surface changes in debris-covered glacier terminus positions. Additionally, InSAR analysis is performed to detect surface changes in the same locations. The manuscript has some relevance presenting new data from a mountain region poorly studied so far. However, the presented study includes too little reference to the state-of knowledge in similar mountain ranges and holds only limited fundamental conclusions of relevance for other parts of the world, or further process understanding. With respect to the scope of the journal, the comments on hazards (GLOFS) remain very general and brief and does not adequately make use of the data presented.
I have significant concerns about the manuscript, as the use of terms, the application of methods and the written presentation of the study require substantial reconsideration and revision before publication. The manuscript presented is also quite brief and could benefit from some more in depth analysis on both lakes and hazards and a thorough language revision and validation.
The most substantial issues are as follows:
A central issue is the definition of lake types and the presentation of different trends and environmental conditions. The authors mapped and analysed lakes in both proglacial and intra-glacial settings. The later are, from my perspective, part of the mentioned, here called “glacier moraine complexes” (GMC). However, for now both lake types are not differentiated and mostly discussed in a mixed fashion. GMC types of terminus conditions are usually termed debris-covered glaciers, or ice-debris complexes and have been studied in many (semi-arid) mountain regions. Formation of lakes within debris-covered ice is referred to as thermokarst, resulting in melting of ice within or underneath the debris cover. These intra-glacier lakes are different compared to proglacial lakes that purely form when ice retreats from an area and water is stored behind dams of debris or bedrock. I think the authors need to reconsider their application and description of the terms and should explicitly recognise these terminological and formative differences in their analysis and presentation. That also includes a better presentation of comparable studies in the introduction. This would provide a more structured and comparable presentation of the data and helps avoiding confusion.
Another issue is the choice of analysis periods. While stated that data from 4 different periods are applied, the results only use 3 periods, summarising the data from 2000-2021. This should be better explained. Why did you not present data from between 2000-2013 (Landsat 7/8)? I suggest including the other periods to better resolve recent trends between 2000 and 2021. I do recognise potential issues due to image availability, but there are ways to get around, as other comparable studies have faced the same issues. Linked to this, the presentation and discussion of trends would improve from a more differentiated temporal analysis. I am convinced that the available remote sensing data can be used for this. Additionally, I miss some analysis of other characteristics of the mapped lakes including more detailed presentation of lake area distribution, changes, dam types etc. for example.
The application of InSAR methods is not adequately presented and seems not to generate any benefit for the study. Surface changes are presented based on DEM-of-Difference analysis and InSAR, but it remains unclear what additional information is generated by the InSAR analysis, since motion data is only referred to as vertical changes. Thus, I think the use of InSAR analysis in this study is mostly obsolete and could be removed entirely.
Finally, the relevance of this topic with respect to hazards (GLOFS) remains very brief and rather general with respect to the scope of the journal. The authors only mention a concentration of population in basins with high lake numbers, but this is only presented in the discussion and not backed by data. No link to previous GLOF events or other reference to an existing GLOF hazard is mentioned in the discussion. This is surely missing for a publication in a journal on natural hazards.
To conclude, this manuscript requires substantial rework and to a part reanalysis. This refers to the issues mentioned above, but also the embedding of this study into a greater scientific framework. Additionally, as mentioned before the language and also the artwork requires substantial improvement before publication. In summary, I guess this ends up in a complete new composition rather than major revision.
Some more detailed comments can be found in the attached pdf and here:
Introduction:
The review of glacier lakes mapping and changes is very limited to the study region, but misses a state-of-the-art overview from the Central Asian environment. There are many studies that address these ongoing changes, also in context of hazards that need to be refered to here to set the study into a broader context. this helps to highlight the scientific significance and surely generates a wider interest in this publication. I suggest to add a more thorough review of the current knowledge on glacier lake dynamics in Central Asia.
Methods:
The methods section is not adequate and should be recomposed to a great extend:
3.1.
The information on remote sensing and DEM data used is limited to Corona imagery currently. Please include more information on all remote sensing products used and also provide all quantitative information available. This helps to document all used data sources and to avoid naming all products multiple times. Please also consider a new, better title for this chapter.
3.2.
The description of the handling of cloud and snow covered area is poorly developed and not convincing. Mapping procedure should also include additional data generated for each lake, e.g. size, river basin name, year of mapping, ice contact or no, lake type (bedrock dammed, moraine dammed). and others…
Furthermore, please give details how you quantify uncertainties given the large variability of resolution in the data used. Please also mention, how you mapped GMCs.
3.3.
This section also has a poor naming. You mention InSAR analysis here, but it is not clear, what they are used for and what kind of geomorphologic analysis is performed and how. You mention basins here and lake basins? What exactly do you refer to here? Further, “surface change of lake basins” is performed using DEMs. What does this mean?
From your figures, I assume you mapped surface depressions in the GMC that may or not contain lakes. I suggest not to use the term basin here, since it sounds too much like river basin. I recommend streamlining the terminology with other studies in ice-debris complexes/debris covered glaciers. These depressions have been termed thermokarst features.
Results:
4.1.
You compare lake sizes between Tien Shan and the eastern Himalayas. Please reconsider your comparison after checking the mapping threshold applied by the other studies. Maybe your difference results from the very small threshold (100 m²) used here.
4.2.
This entire section is very confusing. You mention lakes types in the context of GMCs, but also other lake types. Please rephrase the entire section to put out which lakes are mentioned in which context. I suggest starting with non-GMC glacier lakes and put the special situation of GMCs to the end. One issue on lakes in the GMC, however, is the question of ice content in these areas. You identify “contactless” lake in the GMC, but can you really estimate the ice content underneath the debris? Thus, I question if you really can identify contactless lakes within a DMC area. Please reconsider. Furthermore, the comment on the increase in size of contactless lakes is difficult to understand.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 151 | 41 | 25 | 217 | 7 | 8 |
- HTML: 151
- PDF: 41
- XML: 25
- Total: 217
- BibTeX: 7
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1