the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Feb 2024
A comprehensive rock glacier inventory for Jammu, Kashmir, and Ladakh, western Himalaya, India – Baseline for the permafrost research
Abstract. The prevalent climate warming across the mountain regions worldwide has exacerbated the snow melt, glacier recession and permafrost thawing that is impacting the hydrological cycle. The rock glaciers, a manifestation of ice-rich permafrost, could be regionally important for sustaining the streamflow, especially in the lean season in the Himalaya. Several rock glacier inventories have been developed for high-mountain areas worldwide. However, there are sporadic studies that have characterized rock glaciers in the Himalaya. In this study, a comprehensive rock glacier inventory has been generated for the western Himalayan regions of Jammu-Kashmir and trans-Himalayan Ladakh spread over six mountain ranges utilizing optical satellite images from Google Earth and Sentinel 2A. The inventory has characterized each rock glacier with 22 attributes following the guidelines of the International Permafrost Association (IPA). The inventory contains 5492 rock glaciers (4973 intact and 519 relict) with a total area of 573 km2 and an average area of 0.1 km2. The highest number of rock glaciers (n = 1772) were found in the Zanskar range and the lowest (n = 311) were found in the Pir Panjal range. The majority of rock glaciers (~83 %) have a talus origin, and the remaining ~17 % have a glacier origin. The inventory reports 4756 tongue-shaped and 736 lobate rock glaciers. The average Potential Incoming Solar Radiation (PISR) was observed to be 511 (kWH m-2). The rock glaciers are located between the elevation range of 3301 m asl and 5605 m asl with 63 % of these having a north- or northeast- or northwest-facing aspect. The Mean Annual Air Temperature (MAAT) and precipitation of the rock glaciers range from −8 °C to 8 °C (mean −4 °C) and 71 mm to 1135 mm (mean 328 mm), respectively. The rock glacier inventory provides direct evidence of the presence of permafrost in this ecologically sensitive region and provides a lower bound on the elevation of the permafrost. This inventory shall serve as a baseline for the future hydrological impacts of permafrost and its response to regional climate change. The rock glacier inventory is publicly available at https://doi.org/10.5281/zenodo.10559297 (Bhat et al., 2024).
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(1618 KB)
-
Supplement
(570 KB)
-
This preprint has been withdrawn.
- Preprint
(1618 KB) - Metadata XML
-
Supplement
(570 KB) - BibTeX
- EndNote
Interactive discussion
Status: closed
-
RC1: 'Comment on essd-2023-522', Wilfried Haeberli, 17 Feb 2024
Comments by Wilfried Haeberli
on
A comprehensive rock glacier inventory for Jammu, Kashmir and Ladakh, Western Himalaya, India – Baseline for the permafrost research
Paper submitted to Earth System Science Data by
I.A. Bhat, I. Rashid, A. Banerjee and S. Vijay
General
The submitted article and systematic regional inventory of viscous creep features (rock glaciers) in mountain permafrost has the potential to become a useful contribution to internationally coordinated efforts in support of global climate observation. It must, however, be revised, updated and improved in order to reach modern standards and requirements. This primarily concerns the following central aspects.
As mentioned in the text, the International Permafrost Association (IPA) issues guidelines for compiling standardized rock glacier inventories as a contribution to the Global Terrestrial Network for Permafrost (GTN-P) within the UN/ICSU-related Global Climate Observing System (GCOS). These guidelines have been definitely published in December 2023 (RGIK 2023; https://folia.unifr.ch/global/documents/327247). The contents and guidelines represent up-to-date knowledge and understanding. They are based on consensus of numerous international experts and illustrated with excellent examples. The authors of the submitted paper and inventory for the Western Himalaya should correctly cite this important source and exactly follow the related guidelines.
As part of global climate observation, rock glacier inventories must relate to measured facts about physical conditions, materials, processes and scales in space and time. Intuitive judgements about landform “origins” must be avoided. In this sense, the discrimination between “talus-derived” and “glacier-derived” origins must be replaced by adequate and considerably more differentiated “connection” terms recommended by RGIK (2023), which is explicit on this regard: On page 8 it states that “The spatial connection of the rock glacier to an upslope unit does not necessarily imply a dynamic and/or genetic connection. The term “derived” is not used in this context because it implies an interpretation of the origin of both debris and/or ice.” The long-outdated concept of “talus- and glacier-derived origins” has from the very beginning (Humlum 2010 as cited in the text) been a basic misconception as it compares the ice component in one case (“glacier-derived”) with the rock component (“talus-derived”) in the other.
Concerning conditions, materials, processes and spatio-temporal scales related to the often-spectacular viscous creep features called rock glaciers, the research front is far more advanced than the authors seem to assume. Key aspects are thermal conditions and their change with global warming as documented in borehole temperatures, excess ice formation through slow-deep freezing processes as observed in core drilling and geophysical soundings, geotechnical properties depending on such subsurface ice as analyzed in high-technology laboratory experiments/borehole deformation, and spatio-temporal evolution reconstructed using absolute age dating. As a basis for successful inventory and monitoring work, authors must be familiar with this advanced knowledge basis and use correspondingly precise and up-to-date formulations. A recent “invited perspective” available in “The Cryosphere” at
https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1191/
provides additional help and modern references.
The text can be shortened by focusing on aspects directly related to the presented inventory work and by summarizing quantitative results – especially in section 3.1.1 – by using tables rather than lengthy and repetitive text.
On the technical side: Using the kmz files, I had a look at about 100 randomly selected sites. In view of the difficulty of dealing with small, diffuse or complex features, the impression is one of a large amount of generally useful work. There are nevertheless important open questions, which need careful and critical consideration. The pictures added below illustrate the most important aspects: Is the inventory – as far as possible – really complete and could up-slope connections be treated in a better/more differentiated way? A number of large, clearly recognizable and quite spectacular rock glaciers seem – for unexplained reasons – not to have been inventoried and in a good number of cases, the attribute of up-slope connection is clearly wrong. Concerning up-slope connections, the differentiated RGIK guidelines reach beyond the here-applied simplistic 2-type classification.
Additional comments
More comments can be found in the annotated version of the submitted paper.
Conclusion
Climate-related research on mountain permafrost is an advanced, interdisciplinary research field in full development. In order to reach the required up-to-date quality level, the existing database as well as the submitted article both need critical revision and updating in order to provide a competent and useful contribution to internationally coordinated and highly policy-relevant global climate observation.
Inventory pictures
-
CC1: 'Comment on essd-2023-522', Sheikh Nawaz Ali, 22 Feb 2024
The study entitled “A comprehensive rock glacier inventory for Jammu, Kashmir, and Ladakh, western Himalaya, India - Baseline for the permafrost research” by Bhat et al. indeed provides an exhaustive inventory of rock glaciers in the Northwestern part of Himalaya. However, there are several aspects that need to be improved, revisited, and clarified.
Our comments are attached
-
CC2: 'Comment on essd-2023-522', Ken Hawforth, 16 Mar 2024
While this manuscript addresses a key aspect of high mountain research, i.e., assessing the spatial spread of rock glaciers as potential water resources, it falls short on many key technical points, several of which have already been highlighted by the other reviewer/commentator. The mapping is extremely erroneous as I conclude after assessing over 200 randomly selected outlines, overlaid on Google Earth. Talus/scree slopes, other mass movement deposits, supraglacial debris cover, and terminal moraines have abundantly been marked as rock glaciers. Below are my key remarks:
- Abstract: Calling “rock glaciers, a manifestation of ice-rich permafrost” is fine, but another body of research also links their formation with relict glaciers.
- Abstract: “The inventory contains 5492 rock glaciers (4973 intact and 519 relict)…”. I was keen on learning how the “intact” and “relict” status was decided. Unfortunately, I did not find any conclusive description that suggests any type of geotechnical/geophysical survey performed, or even SAR-based estimations performed in this study to establish even a subset of the identified features as relict or active.
- Abstract: “The majority of rock glaciers (~83%) have a talus origin, and the remaining ~17% have a glacier origin.” Again, how was it determined??
- Abstract: “The average Potential Incoming Solar Radiation (PISR) was observed to be 511 (kWH m-2).” What is the relevance of this information here in the abstract without any further context to this?
- Abstract: “The rock glaciers are located between the elevation range of 3301 m asl and 5605 m asl with 63% of these having a north- or northeast- or northwest-facing aspect.” Again I am not sure what the authors want to signify through this information. There is no further discussion on why this aspect related information is significant if it at all is significant.
- Abstract: “The Mean Annual Air Temperature (MAAT) and precipitation of the rock glaciers range from -8 °C to 8 °C…” MAAT is a single value and not a range!
- I was searching for the available literature on this topic for the study area, and found that the Introduction misses several key references, one of them recently authored by the Commenter Sheikh Nawaz Ali, which provides a rock glacier inventory for nearly the same spatial extent but with largely different statistics than provided by the present manuscript. The inaccuracies in outlines showed by the other reviewers reveal the potential reason behind this discrepancy.
- 1: How was MAAT calculated? Is this for a particular year or an average for several years? The used data must be mentioned in the caption. It is a bit surprising to see a MAAT of -18C at 4,900 m asl and a MAAT of -6C at 5,100 m asl in the adjacent mountain range. This is a huge difference in temperature, certainly defying reported adiabatic lapse rates from this region. There is no description on these in the study area section.
- L122: “However, Google Earth imagery has limited temporal coverage for certain areas which limits the data selection.” But why do we even need high temporal coverage for a rock glacier inventory?! The most important requirement for a RG inventory is having high spatial resolution images to delineate RGs based on the key geomorphic characteristics. 10 m Sentinel 2 images can never be sufficient enough for this purpose. I do not understand the justification behind this data choice. A RG does not suddenly disappear in 1-2 decades without any mark on the landscape, that one needs a high temporal resolution data for inventorying it! This also explains the high inherent inaccuracies in the final inventory.
- L180: Not the best choice of climatic data for this part of the world, and this explains the aforementioned discrepancies in the MAAT values with respect to elevation.
- 3: How is b relict and a intact? What are the visual evidence which enabled this inference?
- L221-222: Again what is the relevance of PISR here? There is no description on this.
- Section 3.1.1: First paragraph again is without any relevance. There is nothing described on how precipitation, MAAT, and RG distribution are interlinked. There is no evidence to suggest how, why, and what has been classified as relict/intact and what was the approach to define its origin (talus vs. glacial).
- L309-319: These are highly speculative descriptions and I fail to understand why the authors are so inclined towards finding any significant correlation between PISR and RGs, when they do not have any in-situ data to support this! Was PISR a decisive factor in the current status of RGs, i.e., intact/relict? A definite evidence-based discussion on this would have been really interesting and enriching.
- Section 4.1: While field work for accuracy estimation of 10 outlines is commendable, it is an extremely small subset of the mapped area. In fact, had highest resolution Google Earth images been consistently used in this mapping, a geomorphologically trained eye could easily spot a RG and its outlines. A better validation process would have been random checks of ~1000 outlines by all the co-authors and then reconciliation of the results. A birds-eye view on a high resolution image is more useful for accuracy estimation of such large datasets.
As such, I find this work inaccurate and incoherent. Unfortunately, an improvement would require a revaluation and mostly regeneration of a large part of this inventory. I do not see it apt for publication in the present form.
Citation: https://doi.org/10.5194/essd-2023-522-CC2 -
RC2: 'Comment on essd-2023-522', Anonymous Referee #2, 17 Mar 2024
The authors present a rock glacier inventory for the Western Himalaya, India. The contribution is of interest to ESSD. However, major problems are identified in terms of paper structure, appropriate selection of cited literature, and most importantly in the methods. The description and documentation of the methods, including: clear statements of the protocol adopted (presently the authors mention an ill-defined integration of Jones et al 2018 and IPA WG RGIK (2018)); clear description of high Sentil-2 were used to integrate GE in areas with poor GE coverage; description of the rules used in the delineation of rock glacier outlines with respect to the rooting zone, which led to inconsistent mapping; description of the morphological attributes adopted in the dynamic classification of rock glaciers (relict and intact). Considering that a similar database requires well-defined and well-documented methods, as well as consistent geomorphological mapping, I recommend a thorough revision of the inventory, before resubmitting a modified version of the accompanying manuscript
1. Introduction
The introduction has a winding structure, which includes a number of repetitions. To streamline the logical flow of section 1, please declare your statement of the problem/s in a more concise and clear way, while ranking the more stringent issues related climate change. Then formulate a clear set of objectives that the compilation of the present inventory conducted on optical imagery (therefore mainly based on qualitative visual interpretation of morphological attributes) can really address. For example, the question of possible mass-wasting related consequences is at least overstated, considering the little contribution (if any) a static RG inventory alone can provide.
Lines 52-54: the distinction between intact and relict features is at least inaccurate. Please start by defining separately intact and relict RGs, then provide the range of attributes that may aid discrimination among them. In this context, please acknowledge the range of uncertainties (widely documented in the literature) associated with this dynamic classification of rock glaciers in inventories compiled via visual interpretation of optical imagery (e.g., Schmid et al., 2015 TC; Brardinoni et al., 2019 ESPL; RGIK, 2023). Similar uncertainties propagate to the range of environmental applications targeted. The authors touch upon uncertainty in section 2.1.4, but the overall question of uncertainty is not sufficiently back up by appropriate referencing to prior work.
Indeed, the selection of the references cited is for the most part, incomplete and not well justified. For example (lines 70-73), the wealth of studies recently conducted on the application of automated and semi-automated mapping efforts is insufficiently covered (e.g., Robson et al., 2020 RSE; Sun et al., 2024 ESSDD). This observation applies to lines 111-114 (section 2.1.1), where the contribution of similar methods is quickly discounted.
Overall, please try to cover, in a more balanced fashion, the range of decades across which the literature has developed (e.g., from the 1950s till today). When citing sample references, please add “e.g.” before the list of relevant studies; when referring to a specific study, please add “i.e.”.
2.1 Study Area
This section should stay separated from “Data and methods”. Please relabel “Study area” as section 2, and “Data and methods” as section 3.
Figure 1. To clearly inform the reader the extent to which the current work builds on (and complements) previous studies conducted in Western Himalaya, it is important that this figure be enriched with the polygon outlines of the study areas covered by existing RG inventories.
Section 2.1.1
Lines 122-129: The authors mention that “Google Earth imagery has limited temporal coverage for certain areas which limits the data selection”, does this mean that for some areas GE imagery was of insufficient quality due to cloud/snow cover? Please clarify, since the compilation of an inventory does not require per se multi-temporal coverage.
Consequently, it is unclear why the authors have decided to complement Google Earth with Sentinel-2 imagery, considering the intrinsically different spatial resolution of these two platforms. Different spatial resolution is going to: (i) set the threshold size beyond which detection becomes unreliable; and (ii) affect the accuracy of RG outline delineation, particularly around the rooting zone. A thorough sensitivity analysis should be conducted, if both platforms are going to be retained in the revised version of the inventory. In line 127, the term “validation” is used, but the entire process remains obscure. Please describe this procedure with further details, providing examples of RGs for which GE imagery was poor and Sentinel-2 was deemed useful. Please add a map, showing which sub-areas have been complemented with Sentinel-2 and which rock glaciers were detected and delineated with Sentinel-2.
Section 2.1.2
Lines 131-134: Inappropriate referencing and reference selection. Please enrich reference to Roer and Nyenhuis (2007) with Barsch (1996) and other classics.
Lines 134-135: “A thorough and comprehensive literature review was conducted to critically evaluate the previous studies on this topic (Table 1).” Please clarify the meaning of this sentence and the practical implication of this review for your work.
Lines 135-136: “In this study, the methodology of Jones et al. (2018) and IPA (Delaloye et al., 2018) was adopted, which offers a solid and organized framework for the analysis of rock glaciers.”
Please replace IPA Delaloye et al (2018) with the actual guidelines included in RGIK (2023). Most importantly, please note that Jones et al (2018) and RGIK (2023 or earlier versions) do not follow the same inventorying rules. Were the two methods integrated? If so, how?
RGIK (2023 or earlier versions) offers baseline and practical concepts for building consistent rock glacier inventories. Some aspects are mandatory, while most of the attributes are optional. When the authors state they have used the methodology proposed by the IPA Working Group on Rock Glacier Inventories and Kinematics (RGIK, 2023 or earlier versions) they should carefully describe which attributes were noted complying with such guidelines. Failing to do so, would make the present inventory incomparable against the current standards and against prospective inventories.
RGIK: Guidelines for inventorying rock glaciers: baseline and practical concepts (version 1.0). IPA Action Group Rock 850 glacier inventories and kinematics, 25 pp, doi:10.51363/unifr.srr.2023.002, 2023.
In Table 1, the following caption is provided: “Comprehensive review and evaluation of previous rock glacier inventories across different mountain ranges of the world”, however, it seems that the list focuses on inventories from continental Asia, rather than worldwide.
Line 143: this is the location where Table 2 should appear and should be referenced for the first time in the manuscript
Lines 145-151: As mentioned by other reviewers/community comments, a real description of how the dynamic classification of RGs in intact and relict is missing. This is a critical gap that needs to be filled.
Lines 153-156: This sentence is a clear example that the RGIK guidelines were not adopted in this study, since therein no distinction between talus-derived and glacier-derived is made. To the opposite, RGIK explicitly avoids making any genetic assumption on RG formation.
Section 2.1.3
Lines 180-181: Please briefly describe what’s the grid size (1 km) and the time window associated with WorldClim 2.1 data. Briefly acknowledge relevant uncertainty and comment on the extent to which inferring climatic controls from such data (MAAT and MAP) should be considered with care.
Figures 2 to 4: all the examples provided show that most of the rooting zone was included. This mapping approach differs from the RGIK guidelines. In the description of your mapping rules, please add the specifics on how the rooting zone was dealt with. If it was included, partly included, or completely excluded. For example, contrary to RGIK mapping guidelines, the rock glacier outline cuts a talus slope in the middle (Figure 2). Please justify this mapping approach. A similar problem applies to Figure 3a).
Section 2.1.4
Subjectivity run over a sample of 20 rock glaciers is not appropriate/informative for an inventory that tallies >5000 of landforms. I suggest either removing this section or to raise the sample size by an order of magnitude.
Sections 4 and 5
Considering the critical issues I have raised on the methodology and mapping examples shown -- such as the need for a clear description of the protocol adopted, the unclear integration between Sentinel-2 and Google Earth imagery, the apparent issues related to inconsistent rock glacier delineation with respect to the rooting zone, unclear rules for dynamic classification – since these are going to directly affect the reliability of the results and relevant discussion, I prefer not to comment on sections 4 and 5.
Citation: https://doi.org/10.5194/essd-2023-522-RC2
Interactive discussion
Status: closed
-
RC1: 'Comment on essd-2023-522', Wilfried Haeberli, 17 Feb 2024
Comments by Wilfried Haeberli
on
A comprehensive rock glacier inventory for Jammu, Kashmir and Ladakh, Western Himalaya, India – Baseline for the permafrost research
Paper submitted to Earth System Science Data by
I.A. Bhat, I. Rashid, A. Banerjee and S. Vijay
General
The submitted article and systematic regional inventory of viscous creep features (rock glaciers) in mountain permafrost has the potential to become a useful contribution to internationally coordinated efforts in support of global climate observation. It must, however, be revised, updated and improved in order to reach modern standards and requirements. This primarily concerns the following central aspects.
As mentioned in the text, the International Permafrost Association (IPA) issues guidelines for compiling standardized rock glacier inventories as a contribution to the Global Terrestrial Network for Permafrost (GTN-P) within the UN/ICSU-related Global Climate Observing System (GCOS). These guidelines have been definitely published in December 2023 (RGIK 2023; https://folia.unifr.ch/global/documents/327247). The contents and guidelines represent up-to-date knowledge and understanding. They are based on consensus of numerous international experts and illustrated with excellent examples. The authors of the submitted paper and inventory for the Western Himalaya should correctly cite this important source and exactly follow the related guidelines.
As part of global climate observation, rock glacier inventories must relate to measured facts about physical conditions, materials, processes and scales in space and time. Intuitive judgements about landform “origins” must be avoided. In this sense, the discrimination between “talus-derived” and “glacier-derived” origins must be replaced by adequate and considerably more differentiated “connection” terms recommended by RGIK (2023), which is explicit on this regard: On page 8 it states that “The spatial connection of the rock glacier to an upslope unit does not necessarily imply a dynamic and/or genetic connection. The term “derived” is not used in this context because it implies an interpretation of the origin of both debris and/or ice.” The long-outdated concept of “talus- and glacier-derived origins” has from the very beginning (Humlum 2010 as cited in the text) been a basic misconception as it compares the ice component in one case (“glacier-derived”) with the rock component (“talus-derived”) in the other.
Concerning conditions, materials, processes and spatio-temporal scales related to the often-spectacular viscous creep features called rock glaciers, the research front is far more advanced than the authors seem to assume. Key aspects are thermal conditions and their change with global warming as documented in borehole temperatures, excess ice formation through slow-deep freezing processes as observed in core drilling and geophysical soundings, geotechnical properties depending on such subsurface ice as analyzed in high-technology laboratory experiments/borehole deformation, and spatio-temporal evolution reconstructed using absolute age dating. As a basis for successful inventory and monitoring work, authors must be familiar with this advanced knowledge basis and use correspondingly precise and up-to-date formulations. A recent “invited perspective” available in “The Cryosphere” at
https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1191/
provides additional help and modern references.
The text can be shortened by focusing on aspects directly related to the presented inventory work and by summarizing quantitative results – especially in section 3.1.1 – by using tables rather than lengthy and repetitive text.
On the technical side: Using the kmz files, I had a look at about 100 randomly selected sites. In view of the difficulty of dealing with small, diffuse or complex features, the impression is one of a large amount of generally useful work. There are nevertheless important open questions, which need careful and critical consideration. The pictures added below illustrate the most important aspects: Is the inventory – as far as possible – really complete and could up-slope connections be treated in a better/more differentiated way? A number of large, clearly recognizable and quite spectacular rock glaciers seem – for unexplained reasons – not to have been inventoried and in a good number of cases, the attribute of up-slope connection is clearly wrong. Concerning up-slope connections, the differentiated RGIK guidelines reach beyond the here-applied simplistic 2-type classification.
Additional comments
More comments can be found in the annotated version of the submitted paper.
Conclusion
Climate-related research on mountain permafrost is an advanced, interdisciplinary research field in full development. In order to reach the required up-to-date quality level, the existing database as well as the submitted article both need critical revision and updating in order to provide a competent and useful contribution to internationally coordinated and highly policy-relevant global climate observation.
Inventory pictures
-
CC1: 'Comment on essd-2023-522', Sheikh Nawaz Ali, 22 Feb 2024
The study entitled “A comprehensive rock glacier inventory for Jammu, Kashmir, and Ladakh, western Himalaya, India - Baseline for the permafrost research” by Bhat et al. indeed provides an exhaustive inventory of rock glaciers in the Northwestern part of Himalaya. However, there are several aspects that need to be improved, revisited, and clarified.
Our comments are attached
-
CC2: 'Comment on essd-2023-522', Ken Hawforth, 16 Mar 2024
While this manuscript addresses a key aspect of high mountain research, i.e., assessing the spatial spread of rock glaciers as potential water resources, it falls short on many key technical points, several of which have already been highlighted by the other reviewer/commentator. The mapping is extremely erroneous as I conclude after assessing over 200 randomly selected outlines, overlaid on Google Earth. Talus/scree slopes, other mass movement deposits, supraglacial debris cover, and terminal moraines have abundantly been marked as rock glaciers. Below are my key remarks:
- Abstract: Calling “rock glaciers, a manifestation of ice-rich permafrost” is fine, but another body of research also links their formation with relict glaciers.
- Abstract: “The inventory contains 5492 rock glaciers (4973 intact and 519 relict)…”. I was keen on learning how the “intact” and “relict” status was decided. Unfortunately, I did not find any conclusive description that suggests any type of geotechnical/geophysical survey performed, or even SAR-based estimations performed in this study to establish even a subset of the identified features as relict or active.
- Abstract: “The majority of rock glaciers (~83%) have a talus origin, and the remaining ~17% have a glacier origin.” Again, how was it determined??
- Abstract: “The average Potential Incoming Solar Radiation (PISR) was observed to be 511 (kWH m-2).” What is the relevance of this information here in the abstract without any further context to this?
- Abstract: “The rock glaciers are located between the elevation range of 3301 m asl and 5605 m asl with 63% of these having a north- or northeast- or northwest-facing aspect.” Again I am not sure what the authors want to signify through this information. There is no further discussion on why this aspect related information is significant if it at all is significant.
- Abstract: “The Mean Annual Air Temperature (MAAT) and precipitation of the rock glaciers range from -8 °C to 8 °C…” MAAT is a single value and not a range!
- I was searching for the available literature on this topic for the study area, and found that the Introduction misses several key references, one of them recently authored by the Commenter Sheikh Nawaz Ali, which provides a rock glacier inventory for nearly the same spatial extent but with largely different statistics than provided by the present manuscript. The inaccuracies in outlines showed by the other reviewers reveal the potential reason behind this discrepancy.
- 1: How was MAAT calculated? Is this for a particular year or an average for several years? The used data must be mentioned in the caption. It is a bit surprising to see a MAAT of -18C at 4,900 m asl and a MAAT of -6C at 5,100 m asl in the adjacent mountain range. This is a huge difference in temperature, certainly defying reported adiabatic lapse rates from this region. There is no description on these in the study area section.
- L122: “However, Google Earth imagery has limited temporal coverage for certain areas which limits the data selection.” But why do we even need high temporal coverage for a rock glacier inventory?! The most important requirement for a RG inventory is having high spatial resolution images to delineate RGs based on the key geomorphic characteristics. 10 m Sentinel 2 images can never be sufficient enough for this purpose. I do not understand the justification behind this data choice. A RG does not suddenly disappear in 1-2 decades without any mark on the landscape, that one needs a high temporal resolution data for inventorying it! This also explains the high inherent inaccuracies in the final inventory.
- L180: Not the best choice of climatic data for this part of the world, and this explains the aforementioned discrepancies in the MAAT values with respect to elevation.
- 3: How is b relict and a intact? What are the visual evidence which enabled this inference?
- L221-222: Again what is the relevance of PISR here? There is no description on this.
- Section 3.1.1: First paragraph again is without any relevance. There is nothing described on how precipitation, MAAT, and RG distribution are interlinked. There is no evidence to suggest how, why, and what has been classified as relict/intact and what was the approach to define its origin (talus vs. glacial).
- L309-319: These are highly speculative descriptions and I fail to understand why the authors are so inclined towards finding any significant correlation between PISR and RGs, when they do not have any in-situ data to support this! Was PISR a decisive factor in the current status of RGs, i.e., intact/relict? A definite evidence-based discussion on this would have been really interesting and enriching.
- Section 4.1: While field work for accuracy estimation of 10 outlines is commendable, it is an extremely small subset of the mapped area. In fact, had highest resolution Google Earth images been consistently used in this mapping, a geomorphologically trained eye could easily spot a RG and its outlines. A better validation process would have been random checks of ~1000 outlines by all the co-authors and then reconciliation of the results. A birds-eye view on a high resolution image is more useful for accuracy estimation of such large datasets.
As such, I find this work inaccurate and incoherent. Unfortunately, an improvement would require a revaluation and mostly regeneration of a large part of this inventory. I do not see it apt for publication in the present form.
Citation: https://doi.org/10.5194/essd-2023-522-CC2 -
RC2: 'Comment on essd-2023-522', Anonymous Referee #2, 17 Mar 2024
The authors present a rock glacier inventory for the Western Himalaya, India. The contribution is of interest to ESSD. However, major problems are identified in terms of paper structure, appropriate selection of cited literature, and most importantly in the methods. The description and documentation of the methods, including: clear statements of the protocol adopted (presently the authors mention an ill-defined integration of Jones et al 2018 and IPA WG RGIK (2018)); clear description of high Sentil-2 were used to integrate GE in areas with poor GE coverage; description of the rules used in the delineation of rock glacier outlines with respect to the rooting zone, which led to inconsistent mapping; description of the morphological attributes adopted in the dynamic classification of rock glaciers (relict and intact). Considering that a similar database requires well-defined and well-documented methods, as well as consistent geomorphological mapping, I recommend a thorough revision of the inventory, before resubmitting a modified version of the accompanying manuscript
1. Introduction
The introduction has a winding structure, which includes a number of repetitions. To streamline the logical flow of section 1, please declare your statement of the problem/s in a more concise and clear way, while ranking the more stringent issues related climate change. Then formulate a clear set of objectives that the compilation of the present inventory conducted on optical imagery (therefore mainly based on qualitative visual interpretation of morphological attributes) can really address. For example, the question of possible mass-wasting related consequences is at least overstated, considering the little contribution (if any) a static RG inventory alone can provide.
Lines 52-54: the distinction between intact and relict features is at least inaccurate. Please start by defining separately intact and relict RGs, then provide the range of attributes that may aid discrimination among them. In this context, please acknowledge the range of uncertainties (widely documented in the literature) associated with this dynamic classification of rock glaciers in inventories compiled via visual interpretation of optical imagery (e.g., Schmid et al., 2015 TC; Brardinoni et al., 2019 ESPL; RGIK, 2023). Similar uncertainties propagate to the range of environmental applications targeted. The authors touch upon uncertainty in section 2.1.4, but the overall question of uncertainty is not sufficiently back up by appropriate referencing to prior work.
Indeed, the selection of the references cited is for the most part, incomplete and not well justified. For example (lines 70-73), the wealth of studies recently conducted on the application of automated and semi-automated mapping efforts is insufficiently covered (e.g., Robson et al., 2020 RSE; Sun et al., 2024 ESSDD). This observation applies to lines 111-114 (section 2.1.1), where the contribution of similar methods is quickly discounted.
Overall, please try to cover, in a more balanced fashion, the range of decades across which the literature has developed (e.g., from the 1950s till today). When citing sample references, please add “e.g.” before the list of relevant studies; when referring to a specific study, please add “i.e.”.
2.1 Study Area
This section should stay separated from “Data and methods”. Please relabel “Study area” as section 2, and “Data and methods” as section 3.
Figure 1. To clearly inform the reader the extent to which the current work builds on (and complements) previous studies conducted in Western Himalaya, it is important that this figure be enriched with the polygon outlines of the study areas covered by existing RG inventories.
Section 2.1.1
Lines 122-129: The authors mention that “Google Earth imagery has limited temporal coverage for certain areas which limits the data selection”, does this mean that for some areas GE imagery was of insufficient quality due to cloud/snow cover? Please clarify, since the compilation of an inventory does not require per se multi-temporal coverage.
Consequently, it is unclear why the authors have decided to complement Google Earth with Sentinel-2 imagery, considering the intrinsically different spatial resolution of these two platforms. Different spatial resolution is going to: (i) set the threshold size beyond which detection becomes unreliable; and (ii) affect the accuracy of RG outline delineation, particularly around the rooting zone. A thorough sensitivity analysis should be conducted, if both platforms are going to be retained in the revised version of the inventory. In line 127, the term “validation” is used, but the entire process remains obscure. Please describe this procedure with further details, providing examples of RGs for which GE imagery was poor and Sentinel-2 was deemed useful. Please add a map, showing which sub-areas have been complemented with Sentinel-2 and which rock glaciers were detected and delineated with Sentinel-2.
Section 2.1.2
Lines 131-134: Inappropriate referencing and reference selection. Please enrich reference to Roer and Nyenhuis (2007) with Barsch (1996) and other classics.
Lines 134-135: “A thorough and comprehensive literature review was conducted to critically evaluate the previous studies on this topic (Table 1).” Please clarify the meaning of this sentence and the practical implication of this review for your work.
Lines 135-136: “In this study, the methodology of Jones et al. (2018) and IPA (Delaloye et al., 2018) was adopted, which offers a solid and organized framework for the analysis of rock glaciers.”
Please replace IPA Delaloye et al (2018) with the actual guidelines included in RGIK (2023). Most importantly, please note that Jones et al (2018) and RGIK (2023 or earlier versions) do not follow the same inventorying rules. Were the two methods integrated? If so, how?
RGIK (2023 or earlier versions) offers baseline and practical concepts for building consistent rock glacier inventories. Some aspects are mandatory, while most of the attributes are optional. When the authors state they have used the methodology proposed by the IPA Working Group on Rock Glacier Inventories and Kinematics (RGIK, 2023 or earlier versions) they should carefully describe which attributes were noted complying with such guidelines. Failing to do so, would make the present inventory incomparable against the current standards and against prospective inventories.
RGIK: Guidelines for inventorying rock glaciers: baseline and practical concepts (version 1.0). IPA Action Group Rock 850 glacier inventories and kinematics, 25 pp, doi:10.51363/unifr.srr.2023.002, 2023.
In Table 1, the following caption is provided: “Comprehensive review and evaluation of previous rock glacier inventories across different mountain ranges of the world”, however, it seems that the list focuses on inventories from continental Asia, rather than worldwide.
Line 143: this is the location where Table 2 should appear and should be referenced for the first time in the manuscript
Lines 145-151: As mentioned by other reviewers/community comments, a real description of how the dynamic classification of RGs in intact and relict is missing. This is a critical gap that needs to be filled.
Lines 153-156: This sentence is a clear example that the RGIK guidelines were not adopted in this study, since therein no distinction between talus-derived and glacier-derived is made. To the opposite, RGIK explicitly avoids making any genetic assumption on RG formation.
Section 2.1.3
Lines 180-181: Please briefly describe what’s the grid size (1 km) and the time window associated with WorldClim 2.1 data. Briefly acknowledge relevant uncertainty and comment on the extent to which inferring climatic controls from such data (MAAT and MAP) should be considered with care.
Figures 2 to 4: all the examples provided show that most of the rooting zone was included. This mapping approach differs from the RGIK guidelines. In the description of your mapping rules, please add the specifics on how the rooting zone was dealt with. If it was included, partly included, or completely excluded. For example, contrary to RGIK mapping guidelines, the rock glacier outline cuts a talus slope in the middle (Figure 2). Please justify this mapping approach. A similar problem applies to Figure 3a).
Section 2.1.4
Subjectivity run over a sample of 20 rock glaciers is not appropriate/informative for an inventory that tallies >5000 of landforms. I suggest either removing this section or to raise the sample size by an order of magnitude.
Sections 4 and 5
Considering the critical issues I have raised on the methodology and mapping examples shown -- such as the need for a clear description of the protocol adopted, the unclear integration between Sentinel-2 and Google Earth imagery, the apparent issues related to inconsistent rock glacier delineation with respect to the rooting zone, unclear rules for dynamic classification – since these are going to directly affect the reliability of the results and relevant discussion, I prefer not to comment on sections 4 and 5.
Citation: https://doi.org/10.5194/essd-2023-522-RC2
Data sets
Rock glacier inventory for Jammu, Kashmir, and Ladakh, western Himalaya, India Imtiyaz Ahmad Bhat, Irfan Rashid, RAAJ Ramsankaran, Argha Banerjee, and Saurabh Vijay https://zenodo.org/doi/10.5281/zenodo.10559297
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 932 | 245 | 46 | 1,223 | 60 | 38 | 34 |
- HTML: 932
- PDF: 245
- XML: 46
- Total: 1,223
- Supplement: 60
- BibTeX: 38
- EndNote: 34
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Imtiyaz Ahmad Bhat
Irfan Rashid
RAAJ Ramsankaran
Argha Banerjee
Saurabh Vijay
This preprint has been withdrawn.
- Preprint
(1618 KB) - Metadata XML
-
Supplement
(570 KB) - BibTeX
- EndNote