A Comprehensive Inventory, Characterization, and Analysis of Rock Glaciers in the Jhelum Basin, Kashmir Himalaya, Using High-Resolution Google Earth Data

Abstract

Rock glaciers are crucial freshwater resources, yet detailed knowledge about their distribution, characteristics, and dynamics in the Himalayan region is scarce. This study presents a comprehensive rock glacier inventory of the Jhelum basin, Kashmir Himalaya, India, using high-resolution Google Earth data. We identified 240 rock glaciers covering an area of 41.24 ± 2.2 km², with ~76% classified as active, ~20% inactive, and 3.7% relict. The average areas and lengths of these rock glacier types were 0.19 km², 0.06 km², and 0.29 km², and 699 m, 426 m, and 952 m, respectively. Most rock glaciers (~90%) were oriented northwards (N, NE, NW), while only 5% faced southwards (S, SE, SW). The lower limit of permafrost in the Jhelum basin is about 3316 m asl. Furthermore, we estimated the ice storage of rock glaciers in the Jhelum basin at 0.80 ± 0.13 km³, equivalent to 0.72 ± 0.12 km³ of water volume. This study enhances our understanding of permafrost distribution and the characteristics and dynamics in the basin. Given their greater resilience to climate change compared to clean glaciers, the hydrological significance of rock glaciers is expected to increase under projected climate change scenarios. This study highlights their importance as a vital water resource amidst the accelerated recession of clean glaciers.

1. Introduction

Snow and glaciers are crucial sources of perennial freshwater [1] for mountain communities worldwide [2,3], especially in dry regions and during dry seasons [2,4,5,6]. High-mountain Asia (HMA), with its numerous glaciers, provides vital freshwater that supports billions of people downstream [7]. Existing glacier studies suggest a rapid retreat of mountain glaciers worldwide [8], largely attributed to anthropogenic activities [9,10]. Over the Himalayan region, enhanced glacier melting with a large intra-regional variation has been reported [11,12]. Recently, Romshoo et al. [13] found that glaciers in the Kashmir Himalaya are retreating faster than those in nearby regions like Nanga Parbat [5], Zanaskar [14], and Ladakh [15,16]. The ongoing retreat and mass loss of glaciers worldwide, especially in HMA, is projected to continue throughout the twenty-first century [9,17,18], posing serious risks to food, water, and energy security [19,20]. Mountain glaciers, excluding those in the Antarctic, are projected to lose about 64% of their volume by the end of the 21st century under the RCP 8.5 emission scenario [21,22], with regions such as Central Europe, Caucasus, Southern Andes, and HMA expected to experience mass losses exceeding 75%. HMA is projected to experience a significant mass loss of ~36% by the end of the 21st century, even if global warming is limited to 1.5 °C above pre-industrial levels [23]. In the short-term, glacier recession leads to increased streamflow; however, once the ‘tipping point’ is reached, melt rates decrease due to glacier shrinkage [24,25], which will cause a gradual decrease in glacier runoff through the end of the 21st century [24]. Most glacierized basins in the HMA region are expected to reach peak streamflow by the mid-21st century under the RCP4.5 climate change scenario [25,26].

Given the projected changes in glaciers, rock glaciers are expected to serve as significant long-term water reservoirs [27,28,29,30] due to their thick supraglacial debris. Unlike clean ice glaciers, they are less susceptible to annual and seasonal variations in meteorological conditions, making them climatically more resilient. Their delayed response to climate change [31,32] further positions rock glaciers as crucial in meeting the increasing water demands in mountainous regions as streamflow from clean ice glaciers diminishes [24,30,33,34,35]. As a result, the hydrological importance of rock glaciers is expected to surpass that of clean ice glaciers in the future [29,33,34].

Despite their hydrological importance, rock glaciers have received relatively little attention globally [7,26,36,37]. This underscores the importance of focusing on rock glaciers, particularly in mountainous regions like the Kashmir Himalaya, where glaciers are receding at a faster rate of 0.75% a⁻¹ compared to the neighboring Ladakh and Karakoram regions [13,38]. Over the past several decades, research has predominantly focused on glacier changes, neglecting other components of the cryosphere like rock glaciers and aufeis [6,39]. Although global attention toward rock glaciers has recently increased, significant data gaps remain in the HMA region. The existing inventories of rock in HMA are sporadic and geographically limited [40,41,42,43,44,45,46], leaving much unknown about their status, distribution, and hydrological significance.

It is well established, as demonstrated in several studies, that the delineation and mapping of debris-free glaciers using automatic and semi-automatic classification techniques on optical remote sensing data are feasible [13,47,48]. However, debris on rock glaciers and debris-covered glaciers are spectrally similar to their surroundings, making it difficult to distinguish these landforms using only spectral information [49]. This similarity limits the effectiveness of traditional approaches for the delineation and mapping of rock glaciers [26]. While previous studies have shown that the synergistic approach combining thermal and optical remote sensing data can be useful for mapping debris-covered glaciers [49,50], this approach is less effective for heavily debris-covered or rock glaciers [51]. Given these limitations, manual delineation remains the most feasible approach for accurately mapping rock glaciers [29,30,40,52,53,54]. Several studies have successfully employed manual delineation techniques, using publicly accessible high-resolution optical satellite imageries such as QuickBird, WorldView, and IKONOS, to compile rock glacier inventories in various regions, including parts of the Himalayas [26,40,54,55,56,57,58,59].

Notably, unlike existing inventories in the Jhelum basin [60,61], this study not only provides a comprehensive rock glacier inventory but also includes a detailed analysis of their water storage capacity. While Majeed et al. [60] inventoried 231 active rock glaciers, we expanded our inventory to include active, inactive, and relict rock glaciers, totaling 240, following the guidelines set by the International Permafrost Association (IPA) Action Group on Rock Glacier Inventory and Kinematics (RGIK). Similarly, Remya et al. [61] focused on 207 active and relict rock glaciers.

The novelty of this work lies in its comprehensive documentation and characterization of all types of rock glaciers, along with estimates of their water storage capacity in the Jhelum basin, Kashmir Himalaya. Furthermore, we conducted a detailed uncertainty assessment of the rock glacier inventory and employed statistical tests, including a correlation analysis and t-tests, to determine the relationship and significance of differences between rock glaciers and their morphological and topographical parameters. The work involves identifying these landforms, understanding their distribution in relation to lithology and topography, examining their development and dynamics, and recognizing their vital hydrological role as potential critical water sources in regions facing glacier recession due to climate change. These findings provide essential baseline data for future research on rock glacier dynamics and their contribution to evolving hydrological systems, as well as support of climate adaptation and sustainable water resource management initiatives in the Jhelum basin.

2. Study Area

The Jhelum basin, whose catchment boundaries coincide with the Kashmir valley, drains into the Jhelum River, a major tributary of the Indus River (Figure 1), covering an area of about 15,534 km² [62]. The basin is flanked by the Pir Panjal mountain range to the southwest and the Greater Himalayan range to the northeast. About 0.46% of the basin area is glacier-covered, with most glaciers concentrated at altitudes between 4500 and 5000 m asl. While the majority of glaciers in the Jhelum basin are clean, a few have considerable supra-glacial debris cover in the ablation zone [63]. As of 2018, the Jhelum basin has 153 clean and debris-covered glaciers, covering ~72 km² [38]. Kolahoi, the largest glacier in the basin [64], is located in the Lidder sub-basin and covers an area of ~11 km². Most of these glaciers are concentrated in the Lidder and Sind sub-basins, covering ~45% of the glacier area in each of the two basins [38]. The climate of the study area is predominantly influenced by western disturbances [65], giving the basin a Mediterranean climate with four distinct seasons: winter (December to February), spring (March to May), summer (June to August), and autumn (September to November). The basin receives an average annual precipitation of ~1200 mm [66], mostly during winters, and primarily in the form of snow. Temperatures in the study area range from −10 to 35 °C [67].

3. Materials and Methods

3.1. Data Sources

The rock glacier inventory was manually created using Google Earth images and complimented by the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Model version 2 (GDEM 02). Google Earth not only allows the exploration of high-resolution satellite images, but also offers user-friendly GIS tools for creating geospatial databases [54,59]. In this study, 210 rock glaciers were mapped using CNES/AIRBUS and MAXAR TECHNOLOGIES optical images from 2017, with a spatial resolution of ~0.5 m, available on the Google Earth platform (

Table 1. Details of image sources and acquisition dates accessed through the Google Earth platform and utilized for rock glacier mapping in the Jhelum basin.

Image Source	Acquisition Date	Rock Glacier Mapped
Maxar Technologies Westminster, CO, USA	4 October 2011	5
CNES/AIRBUS	21 October 2013	7
Maxar Technologies	21 September 2014	9
	22 November 2014
CNES/AIRBUS	9 June 2016	2
CNES/AIRBUS	3 October 2017	210
	6 November 2017
	11 November 2017
	12 November 2017
	29 October 2017
	30 October 2017
	31 October 2017
Maxar Technologies	17 October 2017
Maxar Technologies	14 September 2018	4
	17 October 2020	3

). The platform’s historical imagery navigation tool was particularly useful in identifying rock glaciers, helping to resolve potential ambiguities caused by seasonal snow cover, clouds and shadows [59]. The 3D visualization feature of Google Earth also proved useful in the identification and delineation of rock glacier landforms.

Due to the lack of a high-resolution digital elevation model, we used the relatively coarser resolution ASTER GDEM in this study to compute the topographic parameters of rock glaciers, including elevation, slope, and aspect. The ASTER GDEM V2, with a horizontal resolution of 30 m and a vertical accuracy of 12.26 m [69,70], has been widely used in glaciological studies worldwide [71,72,73,74,75] and was deemed suitable for this study.

3.2. Identification and Mapping of Rock Glaciers

The rock glacier inventory was developed in accordance with the guidelines set by the International Permafrost Association (IPA) Action Group on Rock Glacier Inventories and Kinematics (RGIK) [4]. We divided the study area into 10 × 10 km grids and performed a visual analysis of Google Earth images within each grid to identify the presence of rock glaciers [29,30]. For this, we applied the “geomorphological approach” as outlined in the RGIK initiative [4], which focuses on distinct characteristics of rock glaciers. These include well-developed longitudinal and transversal ridges and furrows indicting signs of flow [76], swollen bodies suggesting the potential presence of ice [77], and steep frontal slopes [77]. The specific geomorphic indicators used as criteria for rock glacier identification and mapping are presented in

Table 2. Geomorphic indicators used as criteria for identifying and mapping rock glaciers in the study area.

Geomorphic Indicator	Intact Rock Glaciers		Relict
	Active	Inactive
Surface flow Structure	Defined furrow and ridge topography [56,76,78,79]	Relatively subdued micro-topography [53,57,79,80]	Less defined furrow and ridge topography [76]
Rock Glacier Body	Swollen body [77] Surface ice exposures[81]	Can have swollen body and surface ice exposures [77,81]	Flattened body [77] Surface collapse features [82,83]
Front Slope	Steep (~>30–35°) [53] Abrupt transition (sharp-crested) to the upper surface [84] Light-colored (little clast weathering) frontal zone, and a darker varnished upper surface [44]	Generally gentler slopes [26,53] Dark-coloured rock-varnished frontal slopes [44,80]	Gently sloping (~<30°) [77] Gentle transition (i.e., round-crested) to the upper surface [84]

.

Following the IPA guidelines [4], we first identified rock glacier landforms, then proceeded to locate each rock glacier and assigned it a unique identifier using latitude and longitude. Additionally, information about the activity of each rock glacier was added to the respective landform in the Google Earth environment. After identification, the rock glaciers were manually digitized as polygons, encompassing the entire rock glacier up to the rooting zone and including the front and lateral margins (extended geomorphological footprint), in accordance with IPA guidelines [4]. The rock glacier polygons, along with basic attributes such as latitude, longitude, ID, and area, were exported to ArcGIS 10.1 software (https://www.esri.in/en-in/products/arcgis-desktop/overview, accessed on 5 June 2024) for further analysis (

Table 3. Description of various rock glacier attributes.

Attribute	Attribute Explanation
ID	Unique identification number (e.g., RG-124)
Mountain Range	[PR] PirPanjal, [GH] Greater Himalayas
Lon	Longitudinal coordinate of polygon centroid (DDD.ddd [N])
Lat	Latitudinal coordinate of polygon centroid (DD.ddd [E])
MEF (m asl)	Minimum elevation at the front
MaxE (m a.s.l.)	Maximum elevation of the feature
MeanE (m a.s.l.) Slope (°)	Mean elevation of the feature Range|Mean
Area (km2)	/
Aspect Class	N, NE, E, SE, S, SW, W, NW (e.g., 90° = E, 180° = S)
Length (m)	Maximum length of the landform
Width (m)	Mean width of the landform
L:W Ratio	Length:width ratio
Geometry	Tongue-shaped, lobate-shaped
Status	Active, inactive, relict

).

In addition to these attributes, the rock glacier inventory included precise and detailed characteristics such as the minimum, maximum, and mean elevation slope, length (parallel to flow), width (perpendicular to flow), and aspect (main aspect of flow), calculated using the ASTER GDEM. The minimum elevation of each rock glacier was calculated at the lowest point where the front slope of the rock glacier met the underlying slope [53]. Similarly, the maximum elevation for each polygon was recorded at the head of the rock glacier. Elevation and slope data were extracted from the ASTER GDEM using the Zonal Statistics raster tool in ArcGIS 10.1 software [29]. Aspect was manually derived along the flow direction and classified into eight classes [59]. Length and width measurements were calculated using the ruler tool in the Google Earth platform [54]. To account for varying widths along the length of a rock glacier, several widths were digitized as lines at intervals of ~50 m perpendicular to the polygon length, and these were then exported to ArcGIS 10.1 software to calculate mean widths (Figure 2; [29]). Furthermore, we used the correlation coefficient [85] to determine the relationships between rock glacier parameters. A t-test was also conducted to assess the significance of the differences observed between various rock glacier parameters [86].

3.3. Classification of Rock Glaciers

Based on geomorphology, rock glacier landforms were categorized as either tongue-shaped or lobate-shaped. A length-to-width ratio greater than 1 indicates a tongue-shaped rock glacier, while a ratio less than 1 indicates a lobate-shaped rock glacier [87,88]. The length and width of a rock glacier are crucial factors in determining its form and contributing area. For instance, a rock glacier that is longer than it is wide is typically constrained by cirque or valley walls, whereas one that is wider than it is long suggests that it receives ice and debris from the sides or walls of the valley [89].

Furthermore, we classified rock glaciers based on their activity. The activity of each rock glacier was determined using geomorphological indicators specified in Barsch [3] and the RGIK initiative [4]. Consequently, rock glaciers in this study were classified into three types: active, inactive, and relict. In the absence of kinematic data, active rock glacier landforms are identified from satellite images as features with a steep frontal slope typically greater than the angle of repose. They exhibited prominent flow-like morphometric features, longitudinal or transverse ridge and furrow assemblages, and sharp-crested front and lateral slopes [4,56,78,79]. Inactive rock glaciers are characterized by geomorphological evidence such as more subdued micro-surface topography and gentler, dark-colored, rock-varnished frontal and lateral slopes (Figure 3; [80]). These landforms may be climatically inactive due to ice melting or dynamically inactive due to the reduced nourishment of talus and/or ice [90].

Relict rock glaciers, which do not show any geomorphological evidence of recent movement, are characterized by collapse structures on their surfaces due to the melting of ice content. They also exhibit subtle micro-reliefs and shallow- and round-crested frontal and lateral slopes (Figure 4; [4,53,79]). Typically, relict rock glaciers have concave longitudinal profiles [82]. Compared to active and inactive rock glaciers, relict rock glaciers may sometimes have vegetated surfaces and are usually situated at lower elevations [53,77]. However, it is noteworthy that the presence of vegetation on a rock glacier is not specifically associated with relict glaciers. Active and inactive rock glaciers are collectively referred to as ‘intact rock glaciers’ [78].

3.4. Rock Glacier Water Storage Assessment

Due to the lack of detailed subsurface data for rock glaciers globally and in the Himalayas, researchers have commonly used empirical thickness–area (H–S) relationships to estimate rock glacier thickness and volume [28,29,30,57,83,91,92,93,94]. The empirical H–S relation used in this study estimates the mean rock glacier thickness $\left(h\right)$ as a function of the rock glacier surface area $\left(S\right)$ in km², using a scaling parameter c = 50 and a scaling exponent $\beta =0.2$, as described below [91]:

$$h=c{S}^{\beta }$$

(1)

The rock glacier volumes are then determined as follows:

$$V=h.S$$

(2)

The water volume equivalent (WVEQ) of rock glaciers was then estimated by multiplying the calculated volume ($V$) by the estimated ice content (% by volume) [26], assuming an ice density of 900 kg m⁻³ [26,95]. While the values of $c$ and $\beta$ are expected to vary across regions [96], existing rock glacier studies typically apply a constant scaling parameter $c$ and scaling exponent $\beta$ when employing empirical power–law relations [28,29,57,91,92,93,94,97,98]. We followed this approach and assumed a standard volumetric rock glacier ice content of 50% [26,91]. Furthermore, our estimation of rock glacier WVEQ included both active and inactive rock glaciers.

3.5. Uncertainty Assessment

While the optical characteristics of rock glaciers are similar to those of debris-covered glaciers [40], mapping rock glaciers presents greater challenges. Visual image interpretation, being inherently subjective, can lead to significant variability in rock glacier outlines, as demonstrated by Brardinoni et al. [99], who observed considerable differences when different analysts mapped the same rock glaciers using high-resolution images. Paul et al. [48] found similar variability in the mapping of debris-covered glaciers. To reduce subjectivity and quantify the uncertainty associated with rock glacier area estimates, a set of 26 glaciers was selected for mapping by eight different analysts. The standard deviation of the area estimates was used to quantify the uncertainty for rock glacier area.

For the accuracy assessment, rock glaciers were randomly selected to ensure representation across all size categories: about 10% in the <0.10 km² and 0.10–0.50 km² size categories, and around 20% in the >0.5 km² size category. Furthermore, this study estimated uncertainty in glacier elevation and slope measurements due to the uncertainties in glacier aerial extents. Figure 5 represents an example of a rock glacier mapped by eight different analysts. The uncertainty in rock glacier ice storage primarily arises from errors in the rock glacier inventory, limitations in the methodology, and inaccuracies in rock glacier thickness estimation. However, this study specifically quantified errors associated with the assumption regarding rock glacier ice content [29], using three values of volumetric rock glacier ice content: 40% (lower), 50% (mean), and 60% (upper), based on previous research [30,57,91,92,94].

4. Results

4.1. Rock Glacier Inventory

In this study, we identified and mapped 240 rock glaciers covering an area of 41.25 ± 2.2 km². The individual rock glacier areas range from 0.003 to 1.03 km², with an average area of 0.17 km² (Supplementary Table S1). Rock glaciers are predominantly found in the Pir Panjal mountain range, with 194 glaciers constituting ~81% of the total in the basin. The Greater Himalayan range, located to the northeast of the study area, contains 46 rock glaciers, accounting for ~19% of the total rock glaciers in the basin. Only two of the rock glaciers exceeded 1 km² in size.

The widths of the rock glaciers range from 55 m to 1330 m, with a mean value of 315 m. The maximum lengths range from 90 m to 2165 m, with a mean length of 654 m. The analysis revealed that 81% of the rock glaciers have a maximum length of less than 1 km, with only two glaciers exceeding 2 km in length. Smaller rock glaciers, with lengths <1 km, exhibited a higher mean altitude (4000 m asl) and minimum altitude at the fronts (MAFs, 3944 m asl) compared to larger rock glaciers over 1 km in length, which had a mean latitude of 3985 m asl and MAF of 3844 m asl. Among the inventoried rock glaciers, only four have widths exceeding 1 km, and 98% of the glaciers have a length less than 1 km, with a mean width of 315 m.

Among the 240 rock glaciers, 124 are smaller than 0.10 km², and 99 fall within the size range of 0.10–0.50 km². Only 17 rock glaciers are larger than 0.5 km², including two that exceed 1 km²; these larger glaciers account for roughly 31.3% of the total rock glacier area in the basin. The rock glacier size category of <0.10 km² covers the smallest area at 13.4%, while the 0.1–0.5 km² category accounts for more than half of the total rock glacier area in the basin (55.2%). Rock glaciers with an area of <0.10 km² are found at a relatively higher mean elevation of 4010 m asl, whereas those in the other two categories have similar mean elevations of 3989 m asl (0.1–0.5 km²) and 3987 m asl (>0.5 km²). Additionally, rock glaciers under 0.10 km² in area have a slightly steeper mean slope of ~21° compared to a mean slope of ~18.5° in the other two size categories (

Table 4. Distribution (count, area) and general topographic characteristics of rock glaciers across various glacier size categories in the study area.

		Area (km2)
Size Category (km2)	Count	Mean	Total	Mean Elevation (m asl)	Mean Slope (°)
<0.10	124 (51.66%)	0.04	5.51 ± 0.3 (13.35%)	4011 ± 198	20.92
0.1–0.5	99 (41.25%)	0.23	22.8 ± 1.2 (55.29%)	3989 ± 194	18.53
>0.5	17 (7.08%)	0.76	12.93 ± 0.6 (31.25%)	3988 ± 192	18.45

).

Furthermore, rock glaciers in the Pir Panjal mountain range are more abundant and slightly larger in size (194 rock glaciers, average size of 0.18 km²) compared to those in the Greater Himalayan range (46 rock glaciers, average size of 0.13 km²). The results of the t-test indicate that the 27.7% difference in average rock glacier size between the two mountain ranges is statistically significant (p = 0.003).

4.2. Rock Glacier Classification

The majority of the rock glaciers (~97%) inventoried in this study are tongue-shaped, while the remaining 3% are lobate-shaped, as indicated by the length-to-width ratio analysis. Furthermore, 183 of the 240 inventoried rock glaciers were classified as active, accounting for 76.25% of the total in the basin, while 20% (count = 48) were classified as inactive. Only nine rock glaciers (3.75%) were classified as relict (

Table 5. Basic information regarding number, area, activity, topographic, and morphological characteristics of rock glaciers mapped in the study area.

		Area (km2)				Elevation (m asl)
Status	Count	Min	Max	Mean	Total	MAF	MaxE	MeanE	Mean Slope (°)	Mean Max Length (m)
Active	183 (76.25%)	0.01	1.04	0.19	35.55 ± 1.8 (86.18%)	3481 ± 145	4485 ± 275	4015 ± 196	19.64	699.44
Inactive	48 (20.0%)	0.00	0.35	0.06	3.1 ± 0.2 (7.52%)	3672 ± 146	4349 ± 276	4010 ± 197	20.99	426.50
Relict	9 (3.75%)	0.02	0.64	0.29	2.59 ± 0.1 (6.28%)	3316 ± 131	3939 ± 247	3650 ± 177	15.71	952.98

). Active rock glaciers cover 35.55 km², accounting for ~86% of the total rock glacier area in the basin. Inactive and relict rock glaciers constitute 7.5% and 6.2% of the total area, respectively. The investigation revealed that the inactive rock glaciers are relatively smaller in size (average of 0.06 km²) compared to active (0.19 km²) and relict (0.29 km²) rock glaciers. The difference in average size among these three types is statistically significant (p = 0.001, t-test). Furthermore, all relict rock glaciers were situated in the Pir Panjal mountain range, while all the rock glaciers in the Greater Himalayan range are inactive.

4.3. Rock Glacier Topography

The inventoried rock glaciers are located at altitudes ranging from ~3300 to ~4500 m asl, with mean minimum and maximum altitudes of 3925 and 4130 m asl, respectively. The majority of these rock glaciers (89.5%) are oriented toward the north, northeast, and northwest, while only 5% are found on southern slopes (S, SE, SW).

Rock glaciers in the study area are slightly more abundant (133) at elevations between 4000 and 4500 m asl (~55.5%) compared to those found at elevations between 3500 and 4000 m asl, where only 107 rock glaciers (~44.5%) were identified. The average size of rock glaciers is 0.19 km² in the 3500–4000 m asl altitude range, compared to 0.15 km² in the 4000–4500 m asl altitude range (Figure 6; Table 5). However, the total area covered by rock glaciers is roughly the same in both the elevation bands.

The mean slope of the rock glaciers in the basin ranges from 9° to 35°, with an average mean slope of 20°. Most of the rock glaciers (183), covering 83% of the total rock glacier area in the basin, have a mean slope between 15° and 25°. The number of rock glaciers with a mean slope in the range of 10–15° (30 rock glaciers) and 25–35° (27 rock glaciers) account for 13.7% and 3.3% of the total rock glacier area, respectively (Figure 7;

Table 6. Distribution of rock glaciers across various elevation and slope categories in the study area. The values in parentheses represent the percentage count and rock glacier coverage within each elevation and slope category.

Elevation (m asl)
		Area (km2)
Category	Count	Min	Max	Mean	Total
3500–4000	107 (44.5%)	0.003	1.04	0.19	20.96 ± 1.1 (50.82%)
4000–4500	133 (55.4%)	0.008	1.03	0.15	20.28 ± 1.1 (49.17%)
Slope (°)
10–15	30 (12.50%)	0.02	0.78	0.19	5.64 ± 0.3 (13.67%))
15–20	104 (43.33%)	0.01	1.04	0.21	22.15 ± 1.2 (58.68%)
20–25	79 (32.92%)	0.01	0.97	0.15	12.12 ± 0.6 (29.37%)
25–30	18 (7.50%)	0.00	0.18	0.06	1.09 ± 0.06 (2.64%)
30–35	9 (3.75%)	0.01	0.06	0.03	0.26 ± 0.01 (0.64%)

). A negative correlation (r = −0.55) was observed between the mean slope and rock glacier area.

Rock glaciers in the study area were predominantly situated on northerly aspects, with 215 glaciers, accounting for 89.5% of the total number of rock glaciers, facing northwest, northeast, and north. Only 12 rock glaciers, accounting for 5% of the total, have southwest, southeast, and south aspects. The remaining three rock glaciers have east or west aspects. Additionally, the average size of rock glaciers on northerly aspects is 0.17 km², which is smaller than the average size of rock glaciers in other aspect classes (

Table 7. Aspect-wise morphological and topographical characteristics of rock glaciers in the study area.

		Elevation (m asl)			Slope (°)			Area (km2)				Length (m)
Aspect	Count	Min	Max	Mean	Min	Max	Mean	Min	Max	Mean	Total	Min	Max	Mean
E	2 (0.83%)	3921 ± 145	4202 ± 274	4062 ± 196	15.50	19.76	17.63	0.44	0.65	0.55	1.10 ± 0.1 (2.65%)	1404.18	1445.78	1424.98
N	215 (89.58%)	3574 ± 145	4301 ± 274	3995 ± 197	9.05	35.49	19.65	0.00	1.04	0.17	35.69 ± 0.6 (86.51%)	90.87	2163.59	630.34
S	12 (5.0%)	3825 ± 149	4315 ± 282	4149 ± 202	13.14	31.17	21.49	0.02	0.97	0.20	2.42 ± 0.1 (5.87%)	134.00	1680.02	718.22
W	11 (4.58%)	3504 ± 141	4177 ± 267	3937 ± 190	14.47	24.18	20.40	0.02	0.39	0.19	2.0 ± 40.1 (2.95%)	281.15	1513.85	914.17

, Figure 8). Rock glaciers with east-facing aspects have the largest mean area of 0.55 km², followed by those on southern and western aspects, with average areas of 0.20 km² and 0.19 km², respectively. However, it is important to note that both rock glaciers greater than 1 km² are north-facing.

Furthermore, it was found that the south-oriented rock glaciers are located at the highest elevations (average 4149 m asl), followed by east-facing rock glaciers, which are located at a mean altitude of 4062 m asl. In contrast, west- and north-facing rock glaciers are located at relatively lower mean elevations of 3937 m asl and 3995 m asl, respectively (Table 7). A significant difference (p = 0.005) in mean elevation was observed between north- and south-facing rock glaciers in this study. Similarly, there is a significant difference (p = 0.001) between south-facing glaciers and those oriented in other directions.

It was found that relict glaciers in the basin are generally situated at lower altitudes compared to the intact rock glaciers, with mean elevations of 3650 m asl and 4010 m asl, respectively. The lowest mean MAF was observed for relict rock glaciers situated at 3504 m asl, while the highest mean MAF was observed for inactive glaciers situated at 3775 m asl. The mean MAF for active rock glaciers, which are situated at altitudes ranging from 3316 to 4200 m asl, is 3575 m asl, which is higher than that of relict glaciers but lower than that of inactive glaciers.

Furthermore, the mean maximum elevation (3820 m asl) of the relict rock glaciers is significantly lower than that of active (4315 m asl) and inactive (4290 m asl) rock glaciers. This is further supported by the t-test, which showed a statistically significant difference (p = 0.001) of 364 m between the mean elevation of relict and intact glaciers in the basin. Additionally, relict rock glaciers were found to have a relatively flatter mean slope (~15.5°) compared to active and inactive rock glaciers in the basin, which have average slopes of 19.5° and ~21°, respectively (Table 5).

4.4. Rock Glacier Water Storage

The rock glaciers in the Jhelum basin have a water storage capacity of 0.80 ± 0.13 km³, equivalent to a water volume of 0.72 ± 0.12 km³, with an average ice thickness of 31 m, ranging from around 15 m to 50 m. Notably, glaciers in the elevation band of 3900–4200 m asl contain the largest WVEQ of 0.55 ± 0.08 km³, representing about 76.3% of the total rock glacier water storage in the basin. Likewise, about 84.7% of the total rock glacier water storage is concentrated within rock glaciers characterized by mean slopes ranging between 15° and 25°. The investigation also revealed that about 87.6% of the total rock glacier ice is stored in north-facing rock glaciers. Furthermore, the Pir Panjal mountain range, which contains the most abundant rock glaciers, stores about 0.62 ± 0.11 km³ of rock glacier water, accounting for 86% of the total rock glacier water storage in the basin.

4.5. Uncertainty Analysis

The uncertainty in the rock glacier area ranges from 0.001 to 0.2 km², with an average uncertainty of 0.03 km², resulting in an overall uncertainty of 19% in the calculation of the rock glacier area in the basin. Variations in delineating glacier outlines between eight analysts were more pronounced toward the upper boundary of the rock glacier landforms (Figure 5). Additionally, high-resolution Google Earth imagery was very effective in differentiating rock glacier landforms from other similar landforms, like heavily debris-covered glaciers, debris flows, and gelifluction, thereby helping in reducing the mapping uncertainty [59].

There is an average uncertainty of 171 m (in the range of 145 to 185 m) and 296 m (in the range of 263–323 m) in the mean MAF and mean maximum elevation, respectively. On average, an uncertainty of 208 m (in the range of 142–225 m) was found in the mean rock glacier elevation. The mapping error is estimated to result in an average uncertainty of 39 m (in the range of 7–165 m) in glacier length and 30 m (in the range of 5–129 m) in glacier width. Similarly, the maximum and mean slope are expected to have uncertainties of 3.2° (in the range of 2° to 5°) and 2.6° (in the range of 1–5°), respectively. However, the general aspect remained consistently the same when the glacier extents were mapped by different analysts. All analysts agreed on the existence of both intact and relict rock glaciers, although there was a disagreement between two of the analysts regarding whether two rock glaciers were active or inactive. Furthermore, we estimated an overall uncertainty of 16.3% in the rock glacier ice and water storage estimates presented in this study.

5. Discussion

5.1. Rock Glacier Inventory and Topographic Distribution

We identified and mapped 240 rock glaciers in this study, including 183 active, 48 inactive, and 9 relict rock glaciers. Due to intrinsic data constraints, such as seasonal snow, cloud cover, and shadowing in the remote sensing data, particularly in the complex Himalayan terrain, the omission of a few more rock glaciers due to human error cannot be ruled out [59]. The rock glaciers investigated in this study are mostly clustered. The average area (0.17 km²), length (654 m), and width (315 m) of the rock glaciers observed in this study are lower than the average area (0.68 km²), length (1600 m), and width (375 m) of rock glaciers reported in the neighboring Himachal Himalaya, India [59]. The higher mean MAF (4484 m asl) of the rock glaciers in Himachal Pradesh, India, explains the presence of relatively larger rock glaciers in that region. Rock glaciers are more prevalent and generally larger in the Pir Panjal range than they are in the Greater Himalayan range of the Jhelum basin. The t-test analysis showed that the 27.7% difference in average rock glacier size between the two mountain ranges is statistically significant (p = 0.003).

The lowest MAF of 3316 m asl observed in this study is consistent with the MAF reported for the western Himalaya (3052 m asl, [59]), Nepalese Himalaya (3225 m asl, [57], Karakoram Hindu Kush region (3500 m asl, [40]), and Tien Shan region (3174 m asl, [100]). The MAF observed in the present study is in good agreement with the lower limit (~3272 m asl) of the permafrost model by Gruber [101] and Gruber et al. [102]. It is worth noting that the MAF for rock glaciers is often considered a good approximation of the lower limit of discontinuous permafrost [53]. This study reveals that no active rock glaciers are situated below 3316 m asl, indicating that this may be the lower limit of permafrost in the Jhelum basin. This is ~884 m lower than the lower limit of permafrost (~4200 m asl) in the Kulu district, Himalcah Pradesh [45] and ~572 m and 684 m lower than the lower limit of permafrost in the Lahul Himalaya (3888 m asl) and the Garhwal Himalaya (4000 m asl), respectively. Similarly, the estimated permafrost lower limit for the Jhelum basin is considerably lower than the permafrost lower limit of 4727 m asl reported by Wani et al. [103] in the Ladakh region. It is also noteworthy that the MAF observed in the present study is slightly higher than the MAF of 3019 m asl reported by Majeed et al. [60] but lower than the MAF of 3791 m asl reported by Remya et al. [61] in the Jhelum basin. Pandey [59] also reported the occurrence of a rock glacier at a lower elevation of 3052 m asl in the neighboring Chenab basin, western Himalaya, which is lower than the minimum elevation of rock glaciers reported in the current study. However, since Pandey [59] identified only a single rock glacier landform, misinterpretation in mapping cannot be ruled out. The presence of a rock glacier below the lower limit of permafrost is due to their slow dynamical downslope motion [45].

We also conducted a comparative analysis of the findings from this study with rock glacier inventories reported from different regions worldwide. However, the analysis is limited to studies that provided data on the number, area, and at least one of the three elevation parameters (i.e., minimum, maximum, or mean) (

Table 8. Characteristics of rock glacier inventories reported from various regions of the world.

Study Area *	Number	Area (km2)	Activity	Min. Elevation (m asl)	Max. Elevation (m asl)	Mean Elevation (m asl)	Source
North America
Colorado Front Range (North America)	220	19.9	Active Inactive Relict	3525 3424 3227	3668 3541 3358	3594 3477 3288	[83]
South America
Central Patagonia	89	14.18	Intact Relict	1766 1758	1941 1919	- -	[104]
Aconcagua River Basin	669	70.0	All	2370	4565	3810	[98]
Central Andes (Atacama region)	477	44.34	All	3807	5504	4427	[105]
Volcán Domuyo region, southernmost Central Andes	224	17.7	Active Inactive Relict	2664 2165 1955	3968 3526 3340	3047 2821 2644	[106]
Monte San Lorenzo massif	177	11.31	Intact Relict	1335 1267	2155 2030	1742 1590	[106]
Valles Calchaquíes Region	488	58.5	Intact Relict	4183 4072	5908 5397	4873 4695	[106]
Huasco Valley, Chile	50	8.6	All	3840	5070	4220	[107]
Europe
Western Tatra Mts.	183	7.14	Intact Relict	1812 1644	- -	- -	[108]
High Tatra Mts.	200	6.7	Intact Relict	2011 1731	- -	-	[108]
Southern Carpathian	306	12.7	All	-	-	1998	[109]
Southern region of the eastern Italian Alps	705	33.3	Intact Relict	2716 1644	3082 2669	2632 2169	[110]
South-eastern Alps	53	3.45	All	-	-	1778	[82]
Tyrolean Alps	3145	167.2	Active	2628	2797	2704	[111]
			Inactive	2542	2665	2598
			Relict	2279	2384	2330
Asia
Nepalese Himalaya	6239	249.83	Intact Relict	4977 4541	5215 4738	- -	[30]
Himachal Himalaya	516	353	All	4484	4900	-	[59]
TienShan of China	261	91.5	All	3174	3486	-	[100]
Daxue Shan, southeastern Tibetan Plateau	295	55.70	All	4352	-	4471	[58]
Central Himalaya	370	28.9	All	4000	6000	5100	[41]
Uttarakhand, Central Himalaya	1004	172.06	All	3821	5822	4742	[112]
Jhelum Basin, Kashmir Himalaya	207	48.23	Active Relict	3591 3573	4461 4588	4026 4057	[61]
Jhelum Basin, Kashmir Himalaya	231	48.46	Active	3019	4633	-	[60]
Jhelum basin Northwestern Himalaya	240	41.24	Active Inactive Relict	3481 3672 3316	4485 4349 3939	4015 4010 2650	Present study

Note(s): * The table contains only those studies that provide data on rock glacier numbers, area, and at least one of the three elevation parameters (i.e., minimum, maximum, or mean).

). From this analysis, we found that the mean minimum elevation of rock glacier landforms in the Jhelum basin is lower than that reported in several other Himalayan regions and higher than that reported in several other mountainous regions globally. As expected, intact rock glaciers are located at higher altitudes compared to relict glaciers worldwide (Table 8). Furthermore, across the Himalayan region, we observed a westward decreasing trend in rock glacier elevation (Table 8). The trend remains consistent when active, inactive, and relict rock glaciers are considered separately. The variations in rock glacier numbers and area in the present study compared to recent inventories by Majeed et al. [60] and Remya et al. [61] can be explained by the fact that this study identified and mapped all three types of rock glacier landforms, unlike the previous studies.

Moreover, the distribution of rock glaciers is climatically controlled by precipitation and temperature, which are considerably influenced by elevation [54]. This explains why the elevation band of 4000–4500 m asl contains more rock glacier landforms than the 3500–4000 m asl elevation band.

The study area exhibits a precipitation gradient of 0.13% m⁻¹ [113] and a mean temperature lapse rate of 7.5 °C km⁻¹ at altitudes above 3100 m asl [62]. The shallow slopes (15–25°) provide a suitable niche for debris accumulation, favoring the development of rock glacier landforms [114]. We found a negative correlation (r = −0.55) between rock glacier coverage and slope. The topographic analysis of our rock glacier inventory showed that 90% of the rock glaciers are north-facing. Due to reduced solar insolation and the correspondingly greater accumulation of snow, ice, and rock debris, north-facing slopes in the Northern Hemisphere are more likely to harbor clean, debris-covered, and rock glaciers [29]. The prevalence of rock glacier landforms on northern slopes found in this study are consistent with other findings in the Hindu Kush Himalaya region [59,100]. This inventory also reveals that, due to reduced insolation, rock glaciers can exist at lower altitudes on northern slopes compared to southern slopes, corroborating previous findings in the Himalayas [53,110]. We observed a statistically significant difference of 155 m (p = 0.001) in the mean altitude between north- and south-facing rock glacier landforms.

Expectedly, relict rock glaciers are situated at lower elevations than intact rock glaciers [77]. The relatively higher temperatures at lower altitudes accelerate ice melting [13,114], increasing the likelihood of relict rock glacier formation. The MAF, or the elevation at the front of an active rock glacier, is often related to the 0°C isotherm [54], which in this study is estimated to be between ~3300 and 4200 m asl. The relict rock glacier landforms (Figure 4), inferred as former rock glaciers, represent a shift in the 0 °C isotherm over time due to climate change and/or debris supply [93].

Notably, the MAF of relict and active rock glaciers showed a mean difference of 113 m, which is thought to represent the upward shift in the isotherm over time [54]. Additionally, future climate change projections for the study area indicate a significant rise in temperatures [113], which may shrink the suitable niches for the formation and persistence of active rock glaciers as the 0 °C isotherm moves closer to or above the maximum altitude of the mountains [54]. As a result of the warming climate, which is expected to be more pronounced at higher altitudes in mountainous regions [115,116], it is predicted that the number of active rock glaciers will continue to decrease while the number of relict rock glaciers will correspondingly increase [111] in the study area.

Furthermore, this study revealed that the Pir Panjal mountain range has a higher abundance of rock glaciers in terms of numbers, coverage, and water storage compared to the Greater Himalayan range. The average rock glacier area of 0.2 km² and 0.1 km² in Pir Panjal and the Greater Himalayan range, respectively, is statistically significant (p = 0.003). Similarly, we observed a statistically significant difference of 91 m (p = 0.0001) in the mean elevation of rock glaciers between these mountain ranges. Glacial history, debris supply, and climate conditions, particularly precipitation and temperature, the latter being significantly influenced by elevation and aspect, play a substantial role in rock glacier development, characterization and distribution [87,117,118]. However, despite more favorable climatic and topographic conditions, such as higher elevation and lower surface temperatures [13,119], the distribution of rock glacier landforms remains limited in the Greater Himalayan mountain range, warranting further discussion and investigation. This suggests that other factors such as rock supply, lithology, and glacial history also impact the formation and persistence of rock glaciers at higher elevations [91,117]. Additionally, the lithology of the two mountain ranges at higher elevations is broadly similar and dominated by volcanic rocks (Panjal volcanics), which are less susceptible to weathering and erosion (Figure 9). The Pir Panjal range, however, is tectonically more active than the Greater Himalayas, uplifting at the rate of 10 mm a⁻¹ [120]. This tectonic activity makes it more susceptible to weathering and debris formation due to the deformation of rocks under compressional stresses, leading to the development of various glacier erosional landforms in the mountain range [121]. Determining the lithological and tectonic control on the rock glacier formation, however, requires more detailed investigation.

Glacier recession and shrinkage are considered to be of great significance for the evolution and origin of rock glaciers, as the transition from alpine glaciers to cirque glaciers often results in rock glacier formation [122,123]. Although the entire study area has been experiencing glacier recession, this phenomenon is more pronounced in the Pir Panjal mountain range [38]. The evidence of former glaciation in the Pir Panjal mountain range is quite apparent in the form of typical glacio-geomorphological landforms [124]. The abundance of rock glaciers in the Pir Panjal range could, therefore, be partly linked to its glacier recession history. In contrast to the 26.2% glacier recession observed in the Greater Himalayan range of the Jhelum basin, the Pir Panjal range has seen a glacier recession of 51.6% between 1980 and 2018 [38].

Figure 9. Lithology of the Jhelum basin encompassing the Kashmir valley, modified after Thakur and Rawat [125].

Figure 9. Lithology of the Jhelum basin encompassing the Kashmir valley, modified after Thakur and Rawat [125].

5.2. Water Storage Estimation

In light of the accelerated glacier recession observed in the study area, the rock glacier water storage (0.72 ± 0.12 km³ WVEQ) will serve as an important water resource at short-, medium-, and long-term timescales. Recent research underscores the importance of rock glaciers as significant long-term water reservoirs, especially in deglaciating and deglaciated semi-arid and arid regions [30]. Consequently, the initial estimates of rock glacier water storage in the Jhelum basin provided in this study are crucial for conducting a more comprehensive assessment of water storage at various timescales. This will enhance our understanding of the dynamics and future implications of these water resources in the context of climate change. The assessment of water resources from rock glaciers will facilitate better management and planning of these depleting resources in the basin, which is currently lacking. Notably, the 157 clean and debris-covered glaciers cover ~72 km² [68] and rock glaciers span 41.2 km² in the Jhelum basin. This highlights their significant role in meeting the increasing water demands for agriculture, horticulture, and domestic use in the basin.

Furthermore, understanding the seasonal variations in rock glacier water flow is crucial for assessing their impact on both ecosystems and human communities [126]. Rock glaciers typically release more water compared to clean and debris-covered glaciers, as the melting process is slower and occurs over extended periods. This delayed response helps buffer the hydrological impacts of seasonal fluctuations, providing a sustained water supply during dry seasons or droughts [127]. Consequently, rock glaciers can play a significant role in ensuring the availability of water resources during peak agricultural demands or dry spells [20,128].

The seasonal meltwater from rock glaciers is also crucial for maintaining ecological balance. Many ecosystems rely on perennial water flows to sustain various plant and animal species [126]. In the Jhelum basin, where seasonal water variability is pronounced during dry periods, the steady meltwater from rock glaciers helps stabilize streamflows, supporting aquatic habitats and riparian vegetation. This buffering effect is important for maintaining biodiversity and ensuring the health of ecosystems that might otherwise be vulnerable to the impacts of a fluctuating water supply [129].

5.3. Impacts of Climate Change on Rock Glaciers

It is pertinent to mention that the temperature in the Jhelum basin is projected to increase by 3 °C and 5.2 °C under the RCP4.5 and RCP8.5 climate change scenarios, respectively, by the late twenty-first century[130]. While precipitation is expected to see a marginal increase by the end of the century, the region is already experiencing significant changes in its form [121,124]. This rise in temperature and shift in precipitation is anticipated to cause substantial long-term changes in the melting and dynamics of rock glaciers, despite their delayed response to climate fluctuations.

In fact, some of the recent studies suggest increased rock glacier surface velocities in response to the warming climate [57]. However, investigations aimed at the assessment of the impact of future climate change on rock glaciers are still in their infancy [26]. Improved modeling efforts focused on understanding changes in the thermal regimes of rock glaciers under various climate change scenarios are necessary. This will enhance our knowledge of rock glacier dynamics and their future evolution, which is crucial for assessing their long-term hydrological impacts.

The rock glaciers’ consistent water supply is invaluable for the Jhelum basin, being crucial for agriculture, horticulture, and domestic use. As ice glaciers recede faster, rock glaciers become increasingly important in meeting water demands. Climate change projections in the Jhelum basin indicate an earlier melt of snow and ice glaciers, shifting streamflow peak to early spring [128,130]. Rock glaciers, with their delayed response to climate change, will ensure stable and perennial water resources. Effective management and planning of these resources are essential to meet the water needs of the basin’s communities and mitigate the effects of diminishing clean and debris-covered glaciers.

Additionally, the investigation revealed that some of the rock glaciers have slid down the valley slopes, even reaching up to roadways and streams, thus threatening lives and infrastructure in the study area (Figure 10). Such rock glaciers have the potential to dam the stream, which upon breach can result in flash floods with catastrophic consequences downstream, particularly under the projected climate change conditions. It is important to note that people are constantly visiting these rock glacier areas for various purposes. The protruding rock glaciers can further slide and obstruct the roadways restricting the movement of tourists, shepherds, and security personnel deployed in these areas. These roadways are used as supply lines by shepherds, cattle herders, and the military, as well as a large number of tourists who trek on these roads round the year.

5.4. Uncertainty

This study provides a comprehensive assessment of the uncertainty in the rock glacier inventory of the Jhelum basin, revealing a 19% uncertainty in the rock glacier area. Analysts showed greater disagreement in delineating the upper parts of rock glacier landforms, with less uncertainty in the termini positions, consistent with previous studies [99]. Since delineating the upper boundary of rock glaciers is challenging [111], this boundary was demarcated based on considered judgement [54,59]. However, the uncertainties associated with delineating upper boundaries are more consistent in this study because the inventory was compiled by a single analyst. Furthermore, the demarcation and mapping of complex rock glaciers, where multiple glaciers merge, followed the methodologies of Scotti et al. [53] and Jones et al. [57], despite their subjectivity [40,53]. The uncertainty in the minimum, maximum, and mean elevations is slightly higher than reported by Brardinoni et al. [99] but comparable to the uncertainty reported by Jones et al. [29]. Additionally, the uncertainty in rock glacier area assessed in this study aligns with the error estimates of Brardinoni et al. [99]. The 16.3% uncertainty for rock glacier ice and water storage is also consistent with previous studies [29,30,57].

6. Conclusions

In this study, we developed a comprehensive inventory of rock glaciers in the Jhelum basin, Kashmir Himalaya, supported with a detailed uncertainty and accuracy assessment of all rock glacier parameters. In comparison to 153 clean ice glaciers occupying ~72 km², this study identified 240 rock glacier landforms in the Jhelum basin, covering 41.24 ± 2.2 km², of which 231 are intact rock glaciers (active and inactive) and the remainder are relict rock glaciers. The majority of these glaciers are located in the Pir Panjal mountain range. Despite favorable climatic and topographic conditions, the limited number of rock glaciers in the Greater Himalayan mountain range suggests an influence of lithology, debris supply, tectonics, and past glaciation on their formation. Additionally, the occurrence of the majority of rock glaciers (90%) on north, northeast, and northwest slopes highlights the role of reduced insolation in their formation and persistence. The absence of active rock glaciers below 3316 m asl suggests that this elevation may be considered as the lowest limit of permafrost in the Jhelum basin, which is lower than the limits reported in several other Himalayan regions.

This study also estimated rock glacier ice storage at 0.80 ± 0.13 km³, equivalent to a water volume of 0.72 ± 0.12 km³, with a predominant contribution from the rock glaciers in the Pir Panjal range. Therefore, this study stands out from previous studies on the subject by incorporating a detailed analysis of the water storage capacity of these landforms in the basin. Due to the limited number of clean ice glaciers, accelerated glacier recession in the Pir Panjal range, and the projected climate warming across the entire study area, the rock glaciers identified in this study are crucial for meeting the region’s water source demands for agriculture, horticulture, industry, domestic use and other uses. The rock glacier inventory and ice storage estimates provide essential baseline data for future research on rock glacier development, dynamics, and their role and importance in the local and regional hydrological regimes of the Kashmir Himalayan region, which is under stress from climate changes.