Glacier dynamics, proglacial lake expansion, and GLOF risk assessment in the lato basin, Trans-Himalayan Ladakh

Abstract

The accelerating retreat of glaciers and expansion of glacial lakes in the Trans-Himalayan region of Ladakh are intensifying the risk of Glacial Lake Outburst Floods (GLOFs), posing significant threats to vulnerable downstream communities. However, high-altitude basins like Lato remain poorly assessed in terms of integrated glacier–lake dynamics and hazard potential. In this study, we present a detailed, multi-parameter assessment of the Lato glacier–lake system over a 40-year period (1980–2020), combining remote sensing, empirical modeling, and hydrodynamic simulations. Our results show an ∼11.5% reduction in glacier area, a mean terminus retreat of ∼171 m at the rate of 4.2 m yr^-1, and a decline in total ice volume from 0.48 km³ to 0.41 km³ (1980–2020). Glacier surface velocities, derived using Glacier Image Velocimetry (GIV), ranged from 0.97 to 15.73 m yr^-1, with mean values between 7.63 and 8.03 m yr^-1. Ice thickness was estimated using GlabTop2 and ensemble modeling approaches, yielding average thicknesses of 76 m and 58 m, respectively. Geodetic mass balance calculations indicate a persistent negative trend, with an average loss of −0.33 m w.e. yr^-1 since the early 2000s. The proglacial Lato lake expanded by ∼66%, reaching 0.088 km² in 2020. Lake depth and volume were estimated using multiple empirical models, with the most robust estimate placing the lake volume at 1.47 × 10⁶ m³. GLOF susceptibility analysis using the Analytical Hierarchy Process (AHP) assigned a high-risk index of 0.82 to the lake. Two-dimensional hydrodynamic simulations using HEC-RAS under full and partial breach scenarios forecast peak discharges of 3726 m³s^-1 and 2309 m³s^-1, respectively, with a potential impact zone extending ∼23 km downstream. This study not only identifies Lato lake as a high-risk GLOF site but also offers region-specific recommendations including glacier and lake monitoring, early warning systems, and the integration of GLOF risk into regional planning and disaster preparedness frameworks.

1 Introduction

Glaciers are crucial water resource for approximately 1.9 billion people globally (Azam et al., 2018; Immerzeel et al., 2020) and significantly contribute to the sea level rise as they melt (Dobhal et al., 2004; Scherler et al., 2011; Hock and Huss, 2021; Bhambri et al., 2018; Hock et al., 2019; Rounce et al., 2023). In recent decades, the Himalayan region has experienced accelerating climatic warming, with a mean rate of increase of 0.21 °C ± 0.08 °C per decade and a maximum rate reaching 0.26 °C ± 0.09 °C per decade, indicating a rising trend over time (Negi et al., 2022). As a result, glaciers showing significant thinning and mass loss at an average rate of 0.3–0.6 m w.e. yr^-1 in the Himalayas (Dehecq et al., 2019; Hugonnet et al., 2021; Zhang et al., 2023). High Mountain Asia (HMA) has witnessed −16.3 ± 3.5 Gt yr^-1 mass loss since 2000–2016 (Brun et al., 2017) accompanied by rate of annual glacier area reduction of 0.04% (Cogley, 2016). This glacier mass loss has resulted in significant ice thinning, which has subsequently impacted glacier velocity and water storage (Dehecq et al., 2019). Approximately 24.5% of the Himalayan glaciers experienced slowdown in velocity (Zhou et al., 2018). The Hindu Kush Himalaya (HKH) region has been experiencing a notable negative trend in glacier velocity, with rates of −1.4 ± 0.4 m yr⁻¹ per decade or −9.8% ± 2.9% per decade (Dehecq et al., 2019). Overall, the combined phenomena of mass balance, glacier slowdown, and ice thinning have led to glacier retreat, contributing to the formation of glacier lakes (Bhardwaj et al., 2025; Ahmed et al., 2022; Singh et al., 2023; Frey et al., 2010; Quincey et al., 2007). In the past half-century, the rapid expansion and formation of new glacial lakes in mountain regions have been primarily attributed to glacier recession (Chen et al., 2024). Unfortunately, this expansion has also significantly increased the risk of substantial hazard in the form GLOFs (Taylor et al., 2023; Sattar et al., 2019a; Shugar et al., 2020; Sattar et al., 2019b; Shea et al., 2015; Singh et al., 2023; Veh et al., 2018; Wang et al., 2015; Westoby MatthewJ. et al., 2014; Westoby et al., 2014 M. J.; Worni et al., 2014; Yao et al., 2018; Zhang et al., 2015; Zhang et al., 2022).

Glaciers in the Ladakh and Zanskar Himalaya show consistent retreat, thinning, and dynamic slowdown. Regional inventories document substantial area loss, including ∼24.9% decline in the Suru Sub-Basin (1992–2023) and sustained shrinkage over four decades (Mankotia et al., 2025; Soheb et al., 2022), while individual glaciers such as Machoi have retreated markedly since 1973 (Taloor et al., 2021). Mass balance studies indicate increasingly negative trends for Pensilungpa and Drang Drung glaciers (Bhattacharjee et al., 2023), supported by recent in situ observations reporting −0.74 ± 0.43 m w.e. yr^-1 for Drang Drung (Azam et al., 2018). Concurrent deceleration and surface lowering in Zanskar glaciers (Ghosh, 2024) further highlight widespread cryospheric change, closely associated with expanding proglacial lakes and rising GLOF susceptibility (Schmidt and Nüsser, 2024; Dubey et al., 2025).

GLOFs are natural hazardous events in which these lakes suddenly release a massive volume of water (Supplementary Figure S1). The mechanism to trigger the GLOFs involves complex process such mass movement impulse waves, mudslide, dam failure, ice calving, sudden snowmelt etc (Richardson and Reynolds, 2000; Anacona et al., 2015; Furian et al., 2022; Taylor et al., 2023; Thayyen et al., 2022; Mishra et al., 2022; Awasthi et al., 2022; Chen et al., 2024). GLOFs can affect downstream areas up to several kilometers, causing damage to the natural environment, socio-ecological systems, human lives, livestock, and infrastructure (Allen et al., 2016; Chen et al., 2024; Carrivick and Tweed, 2013; Singh et al., 2023). Over the past 70 years, the severe impacts of GLOFs have been well-documented, with thousands of fatalities reported worldwide (Emmer et al., 2020). In contrast, only 393 deaths have been recorded in Europe over the past 1,000 years, while the Eastern Himalayas alone have exposed nearly 10,000 people to GLOF-related risks (Carrivick and Tweed, 2016; Kaushik et al., 2024). A study by Shugar et al. (2020) and Abowarda et al. (2021) investigates that glacial lakes larger than 0.05 km² covered an area of 5,930 km² globally between 1990 and 1999. This area increased 1.5 times, reaching 8,950 km² between 2015 and 2018. The number, area and volume of these glacier lakes increase at high rate with increase of 53%, 51% and 48% respectively (Shugar et al., 2020). The spatial distribution and its growth is not homogenous due complex glacier hydrological scenarios such as rapid expansion of glacial lakes was seen at high altitude at Alaska, Greenland and Himalayas whereas small glacial lakes were observed at high altitude in HMA According to Chen et al. (2024), the number of glacial lakes in the Himalayan region has approximately doubled since 1990, consistent with earlier observations reported by Barry (2006). A recent study by Veh et al. (2022) identified 2,800 GLOF events worldwide between 850 and 2022, with the Himalayas showing the highest frequency among global mountain ranges. Over the past four decades, the occurrence rate of GLOFs in this region was estimated at 1.3 events per year (Veh et al., 2019). Using climate models, Furian et al. (2022) predicted that by the late 21st century, approximately 2,700 additional glacial lakes in HMA would trigger GLOF risks. The study also suggested that lake areas will expand by 120%–210%, increasing from 474 ± 121 km² in 2018 to 1,033 ± 148 km² under Shared Socioeconomic Pathway (SSP) scenarios SSP126 and SSP585, respectively. Lake volumes are also projected to rise significantly, reaching 22.8 ± 6.7 km³ under SSP126 and 39.7 ± 7.7 km³ under SSP585—nearly ten times greater than the 2018 estimate of 3.9 km³ (Furian et al., 2022). Zhang et al. (2021) investigates the expansion rates of existing glacial lakes between 2008 and 2016, with maximum rates recorded at 2 myr^-1 in the Himalayas and 1.6 myr^-1 in the Karakoram. These changes were primarily driven by glacier retreat and thinning. According to projections by Linsbauer et al. (2016), a global temperature increase of 1.5 °C (Kraaijenbrink et al., 2017) could lead to the formation of 16,000 new glacial lakes in the Himalayas, collectively holding an estimated total volume of 120 km³.

Remote sensing coupled with field observations is widely employed for glacier lakes studies and inventories (Nie et al., 2017; Zhang et al., 2021). With continued glacier retreat under warming climatic scenarios—especially under high-emission Shared Socioeconomic Pathways (SSPs)—the risk of future glacial lake formation and GLOFs is expected to rise significantly. Using SPOT imagery based on 2° slope threshold, Reynolds (2000) prepared the first inventory of potential future glacial lakes in Bhutan. Subsequent studies utilized satellite imagery and flow models for future glacial lakes identification (Allen et al., 2016; Frey et al., 2010; Linsbauer et al., 2016; Jawak et al., 2018; Zhang et al., 2021; Singh et al., 2023) QUINCEY et al. (2007) incorporated surface glacier velocity to predict future supraglacial lakes, while Frey et al. (2010) applied modelled ice thickness using the Shallow Ice Approximation (SIA) to identify over deepen for lake formation. Similarly, Linsbauer et al. (2016) mapped approximately 16,000 overdeepened sites, representing 5% of the total 28,000 glaciers in the HKH region, which spans about 40,000 km², using the SIA method. Zhang et al. (2021) further employed the SIA approach to locate overdeepened areas in the central Himalayas and Tibetan Plateau, highlighting potential sites for future glacial lake formation. The dynamics of glaciers and glacial lakes in the Trans-Himalayan region of Ladakh, an arid high-altitude zone extremely vulnerable to climate change, have been the focus of many studies (Rather et al., 2024a; Kaushik et al., 2024; Mir et al., 2018; Mir et al., 2018; Thayyen et al., 2022; Immerzeel et al., 2012). These include inventories of glaciers and glacial lakes (Rather et al., 2023a; Majeed et al., 2021; Schmidt et al., 2020), evaluations of lake expansion and glacier retreat, and the wider effects of warming on interactions between glaciers and lakes (Ahmed et al., 2022; Rather et al., 2024b). Significant geomorphic and infrastructure damage was caused by a massive GLOF event that was recorded from the nearby Gya basin in 2014 (Majeed et al., 2021). This event has been thoroughly examined as a warning sign for the region’s increasing cryospheric risks (Majeed et al., 2021; Ahmed et al., 2022; Schmidt et al., 2020). Many sub-basins, like Lato, are still understudied in terms of multi-parameter glacier change and thorough GLOF risk assessment in spite of these efforts.

To address this research gap, we carried out a comprehensive assessment of several glacier parameters and the GLOF risk related to Lato Lake, a glacial lake classified as high-risk. The work combines mass balance assessment, glacial lake dynamics, ice thickness modeling using GlabTop-2 and ensemble techniques, velocity estimation using Glacier Image Velocimetry (GIV), terminus retreat, and analysis of changes in glacier area and length. We also used multi-criteria analysis to evaluate the GLOF susceptibility and two-dimensional hydrodynamic modeling with HEC-RAS to simulate possible outburst scenarios. In addition to identifying key risk factors, we also provide region-specific recommendations for monitoring, mitigation, and policy planning to support disaster preparedness and climate adaptation efforts in Ladakh.

2 Study area

This study focuses on Ladakh’s Lato basin, nestled amidst the Greater Himalayas and Karakoram Range, with altitudes exceeding 3000 m a.s.l (Figure 1). Rain shadow effects create steep precipitation gradients, and sporadic rainstorms, like the 2010 cloudburst, which can cause flooding. The study area spans 33°38′41.884″N to 33°40′15.083″N and 77°33′48.895″E to 77°36′11.091″E, with 3 glaciers (L1, L2 and L3) covering 4.40 km², 2.78 km², and 0.52 km², and snouts at 5251 m, 5291 m, and 5328 m respectively. L1 glacier is associated with a proglacial lake dammed by an unconsolidated moraine with an area of 0.088 km² as per estimates of 2020. Meltwaters from the glaciers are used for meeting various agricultural and domestic needs in the downstream settlement located at around 12.5 km. Lying in the cold-arid Trans-Himalayan region adjacent to Karakoram, Leh (3500 m asl) receives a mean annual precipitation of 100 mm predominantly from westerlies (Chevuturi et al., 2018) indicating dominant input of snow during winter (Rai et al., 2016). Mean monthly temperatures in Leh range from 17.5 °C in August to −7.2 °C in January.

FIGURE 1

3 Methods and datasets

The Landsat time series from 1980 to 2020 with a spatial resolution of 30 m was used for the delineation of boundaries and change detection analysis of Lato group of glaciers and glacial lake (Table 1). Landsat data have been extensively used for decades to investigate the complexity of the earth’s surface dynamics due to its free availability and moderate to high spatial resolution (Coppin et al., 2004; Kennedy et al., 2014). All the Landsat scenes were downloaded from the United States Geological Survey (USGS) web portal http://earthexplorer.usgs.gov/. Additionally, Satellite Radar Topography Mission (SRTM) DEM with spatial resolution 30m downloaded from the web portal www.earthexplorer.usgs.gov Version 3 was used to derive various glacier and glacial lake parameters. The SRTM DEM is widely used for topographic analysis in mountainous regions due to its consistent coverage and reliable elevation data; however, its C-band radar signal can penetrate dry snow to some extent, potentially resulting in slight underestimation of surface elevation in snow-covered areas (Liu et al., 2020). The high-resolution Google Earth (GE) images with very fine resolution ranging between 1.5 and 6.5 m were used to validate the glacier and glacial lake extents derived from the relatively coarser satellite images. Clouds and their shadows hinder the utilization of the satellite scenes. The appropriateness of remote sensing data for the glacier and glacial lake mapping and assessment is determined by the acquisition time, cloud cover, and snow cover (Khan, 2021). To overcome that, end of the ablation season satellite imageries (September) were selected for the present study with less snow, shadow and cloud cover to ensure the accuracy of the results. To estimate glacier velocities, we employed Sentinel-2 and Landsat imagery (2017–2023) using the feature tracking technique (see Supplementary Table S1). Both true colour and false colour composites (Landsat: Bands 5–4–3; Sentinel-2: Bands 8–4–3) were used to enhance ice–terrain contrast and improve tracking accuracy. A total of 31 multi-temporal image pairs were analysed using Glacier Image Velocimetry (GIV) at 30 m spatial resolution. Ice-thickness modelling was performed using the Copernicus DEM (30 m) for the corresponding period. The analysis was based on codes developed by van Wyk de Vries et al. (2022), adapted to meet the specific objectives of this study. Tables 1 and Supplementary Table S1 summarize the datasets and tools used for the Lato glacier group.

3.1 Glacier and lake mapping

Glacier boundaries were mapped using Landsat-9 (15 m) and Sentinel-2 (10 m) imagery following GLIMS guidelines (http://www.glims.org/). Initial delineation was carried out using the Normalized Difference Snow Index (NDSI), a widely applied method for snow and ice detection (Paul and Andreassen, 2009; Paul et al., 2016). The automatically derived outlines were manually refined to correct misclassifications caused by debris cover, terrain shadows, and mixed pixels. Glacial lake boundaries were extracted using the Normalized Difference Water Index (NDWI) (McFeeters, 2013; Bajracharya et al., 2014). Automated results were cross-verified and manually corrected to minimize errors related to shadows and snow cover. Multi-temporal Landsat datasets were used to ensure consistency and capture seasonal and inter-annual variations. Uncertainty in glacier and glacial lake mapping was quantified by incorporating pixel-based positional error, DEM-derived elevation inaccuracies, and manual digitization variability. Glacier outlines extracted using NDSI were refined manually, and an uncertainty of ±1–2 pixels was applied to all mapped boundaries, consistent with established remote-sensing standards for Himalayan glaciers. DEM vertical error (±5–8 m) was propagated into derived parameters such as mean elevation, slope, and aspect. Glacier area and length uncertainties were calculated following the buffer method, resulting in area errors of ±0.02–0.06 km² and length errors of ±20–40 m. For glacial lakes, uncertainties in shoreline delineation introduced similar ±1–2-pixel boundary deviations, producing area uncertainties of ±0.001–0.002 km² and dimensional (length/width) uncertainties of ±5–10 m.

3.2 Glacier surface velocity

Several tools and techniques were employed to derive the surface ice velocity for all the three glaciers.

This includes the widely used technique COSI-Corr involving cross correlation of the optical images, IMGRAFT (Image Georectification and Feature Tracking) and Glacier Image Velocimetry (GIV) toolbox.

3.2.1 COSI-corr

In glaciology, several techniques have been developed and employed to correlate optical images such as cross normalized cross correlation method, correlation operator (CCFO), phase correlation, least square matching, orientation correlation (Heid and Kääb, 2012; Karimi et al., 2012). In this study, we have employed COSI-Corr (plug-in in ENVI software) for quantifying the surface velocity/displacement of the glaciers due to its wide applicability and free availability. The main advantage of using COSI-Corr is that it makes use of single image for each year and this way only two images are needed to derive the displacement, and subsequently the velocity of a glacier between two time periods. One more advantage of the COSI-Corr is it gives both the magnitude as well as the direction of motion of the glacier. For velocity estimation in COSI-Corr Sentinel-2A (10m resolution) images were used in a ENVI based extension plugin. The surface velocity of the Lato group of glaciers was obtained through the cross correlation of the Sentinnel-2A images for the years 2017 and 2023. The mass distribution at the ablation and accumulation area is determined by velocity/displacement of the glacier. In order to lessen the shadow and seasonal snow effects all the satellite data scenes were selected from September and October months. The displacement derived through the cross correlation of satellite images were subsequently checked and corrected.

3.2.2 IMGRAFT

IMGRAFT is a is a MATLAB toolbox developed for analysing displacement within two images between two time periods. Oblique images captured using a simple RGB camera with Ground Control Points (GCPs), aiding in orthorectification as well as satellite images can be used for glacier velocity estimation. This also takes single band images as input. Initially the code was designed for single pair template matching later it was modified to take multiple images as input to reduce uncertainty expecting for better outcomes. Similar dataset of Sentinnel-2A images for the years 2017 and 2023 and additionally Landsat 8 OLI band 8 were also used for the analysis. Apart from the Images, a region of interest (ROI) was delinated for the glacier boundary and correspondingly a mask was used for our area of interest.

3.2.3 GIV

Surface Glacier velocity for Lato group of Glaciers was also derived from feature-tracking-based velocimetry technique involving Glacier Image Velocimetry (GIV), which is an accessible and freely available software package designed for quickly computing glacier-velocity fields at a fine spatial resolution. Feature tracking entails the identification of unique points or characteristics present on the surface of a glacier in consecutive images captured by satellites, drones, or similar platforms. These features may encompass crevasses, variations in ice surface texture, or recognizable landmarks. GIV employs a meticulous process for glacier velocity analysis. Initially, it applies user-specified criteria to filter individual velocity maps, considering both velocity and flow direction. Employing outlier detection functions, it identifies and eliminates anomalous values. Following this, statistical metrics are computed for each grid cell across time, and outliers are flagged by comparing values to the dataset’s mean and standard deviation. Only values within acceptable thresholds for both velocity and flow direction are retained to ensure data integrity. The advantage about GIV is that it uses both the single band as well as RGB images as input for estimating the glacier velocity. In this case, Sentinel-2 images at 10m resolution served as the input data for estimating glacier velocity spanning from 2017 to 2023. GIV was configured to derive surface velocity with a minimum temporal baseline gap of approximately 0.03 Years (11 days approx.) and a maximum of 6 years but the output is in meters per year only. In many cases, image pairs with lag times shorter than 11 days are excluded, as the displacement over such a brief period may be significantly smaller than the offsets caused by distortion or georeferencing errors Van Wyk de Vries et al. (2022). The displacement for each pair of images is computed using a multi-pass frequency domain cross-correlation algorithm. Each pair-wise displacement map is transformed into a velocity map by considering the temporal separation between images. Subsequently, we apply filtering to each map, removing pixels that exceed the maximum velocity threshold, have a signal-to-noise ratio below 5, and exhibit a peak ratio of less than 1.33. Following that, the dataset underwent spatial smoothing for better result and thus removing the outliers.

3.2.4 Area based scaling

Estimation of glacier ice thickness using empirical equations (area-based scaling method) and modeling approaches have been carried out in various glaciated terrains of the world (Farinotti et al., 2009; Morlighem et al., 2011; Paul and Linsbauer, 2012; Sattar et al., 2019a; He et al., 2024; Langhammer et al., 2019).

Where Dg​ is the mean glacier ice thickness (m), and A is the glacier area (km²). The exponent 0.3 represents the scaling factor derived from empirical regression analysis.

3.2.5 Ice velocity based method

The equations used in this study are derived by decomposing glacier surface velocity (Us) into contributions from internal deformation (Ud) and basal sliding (Ub), following the laminar flow principles outlined by Cuffey and Paterson (2010). Assumptions regarding minimal subglacial bed deformation are based on field observations and the framework by Gantayat et al. (2014) and Van Wyk de Vries et al. (2022). This formulation allows direct integration of observed surface velocities into inversion models for thickness estimation.

Here, A denotes the Arrhenius creep parameter, τ represents the basal shear stress at the glacier bed, n is Glen’s flow law exponent, and H signifies the glacier ice thickness.

Here, T is the sole parameter to be estimated, while the constants used are: A₀ = 2.4 × 10⁻²⁴, Qc = 115 kJ mol^-1, the ideal gas constant R = 0.0083145, and the reference temperature T₀ = 273 K. The Shallow Ice Approximation (SIA) method establishes a relationship between basal shear stress and glacier ice thickness, as outlined by Linsbauer et al. (2012) and Gantayat et al. (2014), given by:

In this context, f represents the lateral drag correction factor, defined as the ratio between basal shear stress and driving stress, typically taken as a constant value of 0.8 (Maanya et al., 2016; Gopika et al., 2021). ρ denotes the ice density, assumed to be 900 kg m^-3 as per Farinotti et al. (2009), g is the gravitational acceleration (9.8 m s^-2), and α is the surface slope angle of the glacier. By integrating the aforementioned equations, the glacier ice thickness can be estimated using the laminar flow-based formulation as follows:

H represents the thickness of the ice, U_s denotes the surface velocity, A is a temperature dependent creep parameter and also varying with fabric, grain size, and impurity content. For temperate glaciers, its value is 3.24 × 10⁻²⁴ Pa⁻³s^-1 (Cuffey and Paterson, 2010), f is the slope factor often assumed as 0.8, particularly suitable for Himalayan glaciers (Paterson, 1994), g stands for the acceleration due to gravity (9.8 m/s^-2), α signifies the slope, and ρ indicates the ice density, generally taken as 900 kg/m^-3 (Farinotti et al., 2009).

3.2.6 Basal-shear-stress based method

GlabTop2 is a basal-shear-stress-based slope-dependent gridded ice thickness estimation model for estimating glacier ice thickness and bed topography. It improves upon GlabTop by incorporating spatially variable basal shear stress, using DEM-derived surface slope, a parameterized stress distribution, and an iterative refinement process for greater accuracy across diverse glacier conditions. It is based on the flow mechanics of the average basal shear stress (τ) with shape factor (f) (Nye, 1952; Paterson, 1970).

Frey et al. (2014) provided an empirical relation for computing basal shear stress (τ) along the central flowline, linked to glacier thickness which is of prime importance to derive the glacier thickness (H):

The formula is as follows (Haeberli and Hoelzle, 1995):with f representing the slope factor, typically assumed to be 0.8, while g denotes the acceleration due to gravity (9.8 m/s^-2), α symbolizes the slope, and ρ indicates the ice density, set at 900 kg/m^-3 (Farinotti et al., 2009).

The ice thickness computation is grid-based and requires a glacier mask and DEM as inputs. Initially, a unique ID is assigned to each glacier, and then a code is assigned to all non-glacier cells immediately adjacent to glacier cells. All glacier cells positioned right at the glacier margin are termed as ‘marginal glacier cells’ and they are allocated a unique code. A group of random cells is picked from the inner glacier cells, such that their number corresponds to a preset proportion of the inner glacier cells (Figure 2). Each random cell is then surrounded by an initial buffer of 3 by 3 cells, and each individual buffer is expanded until the difference in elevation between the lowest and highest DEM cells lying inside the buffer is equal to or larger than h_min.

FIGURE 2

The mean surface slope of all glacier cells in the buffer is used to compute an initial estimate of local ice thickness based on Equation 4, and the result is allocated to the random cell to which the buffer was applied. Ice thickness is interpolated to all glacier cells using inverse distance weighting based on ice thickness estimates at all random cells. The value for parameter f is considered as 0.8 which generally works well for Himalayan glaciers (Paterson, 1994). The hmin is considered as 20, and fractions of cells are taken as 30% for random cell pick-up. Here, g represents acceleration due to gravity (9.8 m/s-2), α is slope, and ρ is the ice density, set to 900 kg/m-3 (Farinotti et al., 2009).

In this study, glacier boundaries are obtained from the Randolph Glacier Inventory (RGI), and the DEM is downloaded from https://asf.alaska.edu/data-sets/derived-data-sets/alos-palsar-rtc/alos-palsar-radiometric-terrain-correction/, available at a resolution of 12.5 m. The GlabTop2-py model developed by Wilco-Terink, a Python-based model operating on the PCRaster dynamic modeling framework, is used. The following parameters (Table) were used as input data for the model.

We modeled five different approaches individually and collectively to derive an ensemble mean for glacier thickness estimation. Out of these five models, three methods utilize ice flow velocity based, while the other two rely on basal shear stress. These include VWDV’s Velocity-based inversion with Glaptop2 developed by Van Wyk de Vries et al. (2022), GT2b (Basal-shear-stress-based method in basin-divided approach) Linsbauer et al. (2012), Farinotti et al. (2019), Ramsankaran et al. (2018), GT2w (Basal-shear-stress-based method for the whole glacier), G14b (Glacier ice thickness based on basin-divided approach), and G14w (Velocity-based inversion for the whole glacier). The analysis was conducted using MATLAB codes developed by Van Wyk de Vries et al. (2022) and it was seen that ensemble modelling yield better results than those of individual carried modelling attempts. In the present study we estimated Ice thickness using area-scaling method, and using ensemble modelling as well.

3.3 Glacier volume estimation

Estimates of the amount of ice are essential for evaluating the water reserves stored in glaciers. Though it is not physically feasible to visit each and every glacier so such estimations are particularly very crucial for the Himalayan-Karakoram (HK) area because of its extensive glacier covering (Frey et al., 2014). Volume-area scaling relations have long been used in the past studies for this purpose. The glacier volume can be assessed using three straight forward modelling techniques that utilize data of ice thickness and area. Initially, a piece-wise approach involves summing the product of each ground pixel’s squared size and its corresponding ice thickness. Secondly, the mean ice thickness is multiplied by the glacier area. Thirdly, an empirical approach called volume-area scaling is employed. A comparative assessment of the volume using Empirical formulas as suggested by Bahr et al. (2015) as well as physical based modelling have been applied in the current study to estimate the volume of glacier in the Lato basin. The empirical equation is given as flows:

Where V represents the Volume (m³), A is glacier surface area (km²), and c and γ (dimensionless scaling exponent) are both empirical scaling parameters with values 0.191 (km³/km²^γ) and 1.375 respectively.

We also adopted a piece-wise (pixel-wise) approach for estimating glacier ice thickness and volume. In this method, the glacier domain was discretized into individual raster pixels, and the ice volume was calculated by multiplying the ice thickness of each pixel by its corresponding pixel area and subsequently summing over all glacier pixels. All volume computations were carried out using MATLAB. Using the piece-wise approach, the estimated glacier volume is 2.25 km³. For comparison, the glacier volume was also calculated using the volume–area scaling approach, which yields a total volume of 2.69 km³.

Here, V_m denotes the total glacier ice volume (expressed in km³), A_p represents the area of an individual raster pixel (m²), and h_i​ is the ice thickness corresponding to the i-th pixel (m). The variable N indicates the total number of pixels constituting the glacier domain.

3.4 Topographical characteristics

The distribution and dynamics of glaciers are significantly influenced by crucial topographical attributes such as elevation, aspect, slope, and hypsometry. These parameters play a vital role in determining the behavior and spatial extent of glaciers. Moreover, the processes of ablation and accumulation are largely controlled and determined by topographic structure of glaciers. In this study attributes like Area, Length and Perimeter were directly extracted by using glacier polygons (DeBeer and Sharp, 2009). Other glacial parameters like Elevation, Slope, Aspect, Hypsometry of each glacier were calculated by using high resolution ALOS PALSAR Digital Elevation Model (DEM) and derived Glacier polygons. The whole process was carried out in GIS environment using the concept of zonal statistics.

3.4.1 Equilibrium line altitude and geodetic glacier mass balance

Mass balance is one of the important and essential parameters for assessing the sensitivity of glaciers towards climate change (Tawde et al., 2019). Mass balance of any glacier is said to be negative if melting is more than accumulation and positive if accumulation exceeds melting. There are different direct and indirect techniques for estimating mass balance of glaciers. However, direct mass balance measurements are not feasible for every glacier owing to their remote geographic location and limited resources. Optical remote sensing sensors offers a huge potential for estimating mass balance of individual glaciers remotely. In this study, mass balance of glaciers was estimated by using accumulation area ratio (AAR) method which requires the positional knowledge of ELA. ELA is an imaginary line on a glacier where net accumulation equals to net ablation during hydrological year (Tak and Keshari, 2020). Since the glaciers of study region are temperate glaciers therefore, satellite-based end of the ablation season snow lines was taken as representatives for the Equilibrium Line Altitude (Paterson, 1998). ELA of all three glaciers in Lato basin was delineated manually using different band combinations of high-resolution sentinel-2 imagery of 2020 and digital elevation model. The ELA was used to calculate AAR of glaciers, which is the ratio between accumulation area and total area of glaciers (Meier, 1962). Finally, AAR values were incorporated in the regression equation developed by Tawde et al., 2019 (Equation 1) for Chandra basin Glaciers to estimate mass balance of each glacier.

Where B is annual specific mass balance (cm w.e)

The glaciological mass balance method which needs multiple field-based observations including the installation of the ablation stakes on the surface of the glacier was not possible for the present study because it requires enormous financial support and manpower. The mass balance of glacier was obtained using the geodetic technique. The geodetic mass balance corresponds to glacier-wide volume changes, converted to mass changes by applying density assumptions (Hugonnet et al., 2021; Van Tricht et al., 2023). This process typically involves the comparison of multi-temporal digital elevation models (DEMs), requiring accurate and repeated mapping of the glacier’s surface topography. It is assumed that bedrock elevations remain constant, implying that surface elevation changes correspond to variations in ice thickness. By integrating the observed surface elevation changes across the entire glacier, ice volume changes are determined. Subsequently, these volume changes are converted into mass changes, incorporating density assumptions (Huss, 2013). Geodetic mass balances serve as a primary tool to retrieve changes in glacier mass at local, regional, and global scales (e.g., Brun et al., 2017; Shean et al., 2020; Bhattacharya et al., 2021; Hugonnet et al., 2021). Beyond this primary application, geodetic mass balances play a crucial role in the calibration and correction of cumulative mass balances derived from glaciological methods (Zemp et al., 2013) and in the calibration of mass-balance models (Huss and Hock, 2015).

3.5 Glacial lake depth, volume and peak discharge estimation

The surface area of a glacier lake becomes an important indicator for measuring lake volume and hazard because the depth of a glacial lake is difficult to determine using remote sensing data and GIS, in such cases empirical equations developed in the different parts of the world are used (see Supplementary Table S2). Estimation of parameters such as depth, volume and potential peak discharge is essential step prior to the assessment and simulation of a GLOF event. There are various equations available in the literature to estimate such parameters. Here we used the widely employed empirical formula given by Huggel et al., 2003 to calculate the lake depth and were subsequently compared with other formulas given by Fujita et al., 2013; Patel et al., 2017, whereas volume was calculated by several formulas generated for different regions of Himalayas such as for Koshi basin of Nepal Himalayas by Liu et al., Skim Himalayas by Sharma and Shukla (2018), Himalayan region by Fujita et al., 2013, Chandra basin western Himalayas by Patel et al., 2017. Besides that, empirical equations developed for regions outside Himalayas were also included such as for Swiss Alps by Huggel et al., 2002, Cordillera of western Canada by Evans, 1986. The results obtained by employing all these equations showed a considerable variation. Since current volume-area scaling relationships sometimes obscure the intricacy of lake geometry, we used the improved equation of volume estimation specific to moraine dammed lakes developed by Qi et al., 2022 obtained from the large lake bathymetry datasets. The present volume area scaling relationship is based on the assumption that variables/parameters such as volume, area and width/length ratio scale predictability (Qi et al., 2022). The effectiveness of the newly developed equation has been compared with the already-established empirical equations, overall, the newly developed volume-scaling relationship developed by Qi et al., 2022 provides approximate mean values than other previously reported formulae. The equation is given as below:where V was the lake volume (×10⁶ m³), A was the lake area (km²), and Ratio (mxw/mxl) was the maximum width divided by the maximum length of the glacial lake.

Similarly average peak discharge was calculated using three empirical equations given by Evans (1986), Popov (1991), Huggel et al. (2002), subsequently an average of these equations has been used in the present study as suggested by Emmer et al. (2018), Chowdhury et al. (2021). The details of the equations used to calculate the depth, volume and peak discharge is given in Supplementary Table S2.

3.6 GLOF susceptibility

In order to assess the GLOF Susceptibility of the lake, a number of parameters/indices (lake area, volume, elevation, dam condition, presence of lake in the upstream/downstream region, lake expansion, glacier recession, distance between lake and mother glacier, distance between lake outlet and nearest settlement or infrastructure, probable rockfall, ice or snow avalanche zone and slope between the lake and mother glacier) have been reported in the past studies. However, there is no set criteria mentioned in the scholarly literature, which can be used for the selection of the parameters for GLOF susceptibility (Ahmad et al., 2022). Field investigations are recommended to accurately derive the GLOF susceptibility parameters, but field survey is not possible for all the lakes due to their remote locations and hilly terrain. Therefore, geospatial techniques play a key role to derive such parameters in absence of the field observations and it has been widely applied in the past studies very efficiently to assess the GLOF susceptibility. In the present study, seven important parameters (see Table 2) have been selected for the GLOF susceptibility of the Lato glacial lake based on their level of importance. The details of the parameters used for the GLOF susceptibility are given in Table 2. The threshold of the parameters has been derived from the past studies. SRTM DEM, Landsat imagery, Google earth imagery have been utilized to derive these parameters. Parameters such as glacier area, glacial lake area, glacier and lake dynamics, distance between lake and glacier, elevation have been calculated through the multi-temporal satellite imagery and DEM. Presence of potential ice and snow avalanche was visually checked on high resolution GE imagery.

After the identification and calculation of parameters weights were allocated to each factor using Analytical Hierarchy Process (AHP). AHP model propounded by Saaty, 1980 has been widely used for the hazard and risk assessment such as land degradation, landslides, snow avalanche, debris flow (Shahabi et al., 2015) and GLOF hazard assessment (Khadka et al., 2022). Each parameter/factor were ranked into three classes viz High, Medium and Low with a corresponding Index value of 1, 0.25 and 0.5 (Prakash and Nagrajan, 2017). The final weights for each of the parameters/factors were determined by multiplying the factor weight (Wi) by the class index values (Ci) based on the measurement of lake factors. At the last stage, final weights of each factor were added to assess the GLOF Susceptibility of the Lake (Equation 11).

The GLOF susceptibility index value ranges from 0.25 to 1. Using an equal interval class system, the values are divided into five categories: Very Low (0.25–0.40), Low (0.40–0.55), Medium (0.55–0.70), High (0.70–0.85), and Very High (8.5).

3.6.1 Simulations of GLOF using HEC-RAS

Flood water movement through any natural landscape resulted due to the failure of the dam can be predicted using various kinds of methods. A dam breach can pose a serious threat to the people and infrastructure downstream. Such events are uncertain; thus, they can breach unexpectedly. Humans can’t prevent or stop such catastrophic events. However, assessment and subsequent Modeling can predict the Flood water movement. Although modeling can’t be properly accurate, still they can provide us a better idea of the extent up to which flood water gets inundated at the time of the Dam-Break event. Describing natural physical phenomena using numerical methods requires making broad assumptions to develop governing equations. While simple hydraulic modeling methods can be adequate for the approximate propagation of flood peaks through river channels. Advanced models are capable of modeling more detailed physical phenomena, but this does not correspond to a decrease in uncertainty. The consequences of the Dam failure can be disastrous both in terms of human loss and the destruction to the property and infrastructure downstream. Hydraulic models offer an opportunity to provide a picture of the probable consequences of partial or complete dam failure, which is crucial to carry out emergency planning and conducting reliable risk assessment. Hydraulic modeling is an essential component of establishing a strong flood forecasting framework. Inundation maps, which are believed to be most useful outcomes produced from flood simulations can be prepared by the resulted simulations from the hydraulic models, which may be used by the officials to evaluate the risk of flood.

This study assesses the potential GLOF hazard associated with Lato lake using a combination of remote sensing analysis, hydrodynamic modeling, and sensitivity assessment. The HEC-RAS hydrodynamic model was used to simulate potential GLOF scenarios under varying breach conditions. We analyzed the impact of GLOF on the downstream area of 23.14 km through two-dimensional unsteady flow analysis in HEC-RAS. This model analysed key parameters such as peak discharge, flow velocity, and inundation depth were analyzed for two villages: Lato (immediately downstream) and Miru (further downstream).

The lake volume is calculated based on Huggel’s empirical formula:

Where V represents the volume in cubic meters, and A is the lake area in square meters.

The volume of Lato lake was estimated to be 1.046 × 10⁶ m³ using Huggel’s method.

3.6.2 Scenario analysis

4 Results and discussion

4.1 Characteristics of glaciers in lato basin

In this study, we examined three glaciers coded as L1, L2, and L3. All these glaciers are winter-accumulation type and partially debris-covered. Based on their morphology, L1 and L2 are compound-basin valley glaciers, whereas L3 represents a simple-basin valley glacier. The total glacierized area is 7.74 ± 0.05 km², with L1 covering 4.40 ± 0.03 km², L2 covering 2.80 ± 0.02 km², and L3 covering 0.50 ± 0.01 km². These uncertainties correspond to typical mapping errors (1–2 pixels) associated with NDSI extraction and manual refinement (Paul et al., 2013; Bolch et al., 2012). The glaciers lie between 5250 ± 5 m and 5850 ± 5 m, with a mean elevation of 5591 ± 6 m, reflecting DEM vertical uncertainty reported in Himalayan regions (Berthier et al., 2006; Mukherjee et al., 2018). The mean slopes are 12.30° ± 0.3° (L1), 12.52° ± 0.3° (L2), and 25.60° ± 0.5° (L3), consistent with slope error values in 30 m DEM products. The mean glacier length is 2800 m, ranging from 1684 m to 3796 m. Length uncertainty mainly arises from pixel size and terminus digitization (Paul et al., 2013). The estimated ice reserves in the three glaciers are 0.40 ± 0.01 km³, where uncertainty incorporates error propagation from area and thickness models (Farinotti et al., 2019).

4.2 Glacier area and length changes

The datasets generated from multi-temporal and multi-sensor satellite imagery reveal noticeable changes in glacier dynamics from 1980 to 2020 in the Lato glacier group. The total glacier area decreased from 8.72 ± 0.06 km² in 1980 to 7.70 ± 0.05 km² in 2020 (Figure 3), indicating a net loss of ∼1.02 ± 0.08 km² (≈11.5%) over the 40-year period, corresponding to an average retreat rate of 0.025 ± 0.002 km² yr^-1. These uncertainty values reflect typical mapping errors of ±1–2 pixels in glacier delineation derived from NDSI and manual correction (Paul et al., 2013; Bolch et al., 2012). Area loss was 6.6% during the first 2 decades and 4.9% during the last 2 decades. Among the glaciers, L3 exhibited the highest relative loss (∼17.15%), while L2 and L1 showed moderate losses of ∼12.3% and ∼9.7%, respectively. The higher relative loss of L3 may be attributed to its smaller size, as smaller glaciers generally respond more rapidly to climatic forcing due to shorter response times and limited accumulation zones, whereas the comparatively higher mean elevation of L1 likely moderated its retreat. Glacier length also decreased, with the mean length reducing from 2960 m in 1980–2800 m in 2020. L1 showed the maximum total retreat (183 m), while L2 and L3 exhibited average annual decreases of ∼3.9 m and ∼3.5 m, respectively. Like mass balance, the terminus position of glaciers and its fluctuations with time is an important parameter to assess the health of glaciers. Together, the glaciers have shown a mean shift of around 171 m at the rate of 4.2 m a-1 in their terminus position during these 4 decades (Figure 3). With an annual shift of 5.8 m, L1 glacier has shown a maximum upward shift followed by L3 (4 m a-1) and L2 (3.12 m a-1).

FIGURE 3

Area altitude distribution of glacier area was calculated by classifying the DEM data of glaciers into 100-m equal bins and further calculating the area of each elevation zone by converting it into vector format. Mean elevation of glaciers was found to be 5591 masl, ranging from 5559 (L2) to 5618 (L1) masl. Further, hypsometric curve of glaciers revealed that almost 76 percent of glacial area is concentrated within the elevation range of 5450–5650 masl (Figure 4).

FIGURE 4

4.3 Glacier mass balance (AAR method)

Considering end of the ablation season transient snow line (TSLs) of glaciers as equivalent of Equilibrium Line Altitude, the mean ELA of glaciers was demarcated at an elevation of 5590 masl. Using hypsometry of glaciers, the total accumulation area and accumulation area ratio (AAR) of glaciers was estimated to be 3.51 km² and 0.44 respectively (Table 3). By applying the individual values of AAR in the regression developed by Tawde and others for estimating mass balance of glaciers in the Chandra basin, the mean mass balance of glaciers was calculated to −40.82 cm.we. All the three glaciers are showing the negative mass balance. The specific mass balance of the glaciers varies from lowest value of −35.2 cm w.e (L1) to a high value of −44.55 cm w.e. (L2).

4.4 Glacier mass balance (geodetic method)

Geodetic mass-balances can be particularly useful in regions with sparse direct glacier observations, like the Lato group of glaciers in Central Asia, where accessibility is poor and climate conditions are harsh. Here, we analyse the geodetic mass balance for the group of glaciers as determined from two different estimates (Brun et al., 2017; Hugonnet et al., 2021), and we reconstruct the geodetic mass balance using the TanDEM-X Change DEM (Lachaise et al., 2021), which provides elevation changes between 2012 and 2019 (Figure 5). For the latter, a volume-to-mass conversion factor of 850 kg m^-3 is used, similar to Brun et al. (2017) and Hugonnet et al. (2021). It must be noted that the TanDEM-X DEMs used to infer the elevation changes were created based on data obtained in winter. Because the synthetic aperture radar (SAR) data can be biased by differences in signal penetration depth into the glacier surface, this might have led to an underestimation of the elevation changes (Li et al., 2021). We therefore append a relatively large uncertainty value of ±0.2 m w.e. yr^-1 to the determined geodetic mass balance, covering the possible underestimation. Combining the existing geodetic mass balance estimates and the one derived from the TanDEM-X DEMs, it can be observed that these were for all periods (largely) negative (Figure 6). The average over the first 2 decades of the 21st century was in the order of −0.33 m w.e. yr^-1, indicating a strong mass loss over time. This shows that the Lato group of glaciers has been losing mass over time. The most eastern glacier, RGI14.14436 seems to lose mass at a decreasing rate. This might be caused by this small glacier to rapidly retreat to higher elevations (as it is very small) making it less sensitive to climate change.

FIGURE 5

Lachaise et al., 2021).

FIGURE 6

Brun et al., 2017; Hugonnet et al., 2021) and the one derived in this study (solid blue line). The dashed lines show he intermediate geodetic mass balance for the three individual glaciers based on Hugonnet et al. (2021). The solid lines show the weighted average for the three glaciers.

4.5 Spatial distribution of ice thickness

4.5.1 Using GlabTop model version 2

The GlabTop model version 2 was applied for the RGI glacier boundaries using the parameters mentioned in the Table 4. The analysis revealed that, ice thickness of the Lato basin glaciers (L1, L2 and L3) varied between 7m and 165m with a mean thickness of 76 m for all the three glaciers. The minimum and maximum ice thickness for the L1 glacier was estimated to be 13.59 m and 164.35 m respectively, whereas mean thickness was calculated as 74.65 m. The spatial distribution of ice thickness was also analysed by reclassifying the ice thickness map into 4 classes at equal intervals to estimate the area under each thickness class. As mentioned in the Supplementary Table S3, maximum area 43.60% (2.20 km²) of the L1 glacier area comes under the ice thickness class 2 (51–100m), 29% area is covered with ice thickness 0–50 m, whereas only 0.06% area of the glacier is covered with the ice thickness greater than 150m (Figure 7a). In the L2 glacier of Lato basin, maximum and minimum ice thickness varies between 9.2 m and 138 m, while mean ice thickness was estimated as 58.94 m. The maximum area (45.18%) of the L2 glacier is covered with the thickness class 1 (0–50 m) reported in the Supplementary Table S3, followed by the thickness class 2 (50–100 m) covering an area of 1.64 km², which is 42% of the total glacier (L2) area (Figure 7b).

FIGURE 7

The minimum and maximum ice thickness for the L3 glacier was calculated as 7.63 m and 84.23 m respectively with a mean thickness of 34.31 m. As shown in the Table 3 the 82% of the glacier (L3) area is covered between the thickness class of 0–50 m, whereas 17% of the area comes under the thickness class 51–100 m (Figure 7c). The maximum ice thickness was observed in the central parts (central flow lines) and near the ELA of the glacier, whereas lowest thickness (pixel value) was observed at the glacier boundaries. Gantayat et al. (2014) have estimated the ice thickness of Gangotri glacier using GPR show that ice-thickness of the glacier varies from 15–540 m with maximum thickness in central part and minimum near the snout of the glacier was observed. The similar phenomena was observed by Ramsankaran et al. (2018) for the Chota Shigri glacier.

Our finding on ice thickness for glacier L3 with an area of 0.72 km² are in agreement with the past study carried out by Majeed et al. (2021) in the adjacent basin of Lato watershed using GlabTop version 1 which shows the ice thickness of the Gaya glacier (0.50 km²) ranging between 1–55 m with a mean thickness of 28 m. The presence of highest ice thickness in the L1 glacier is probably due to the lower rate of area retreat (17% in past 48 years), whereas it is high in case of Gaya glacier (46% in past 50 years) and L3 glacier. Ice thickness of the individual or a group of glaciers has been estimated extensively in various glacierised basins using empirical equations and modeling approaches for example, Gharehchahi et al. (2020) have estimated the ice thickness of glaciers in Upper Rhone Catchment using GlabTop model. Linsbauer et al. (2016) estimated ice thickness and possible future lakes for glaciers in Karakorum Himalayas. The ice thickness of 65 glaciers estimated through GlabTop-2 model in the Chandra basin reveals that ice thickness of all the glaciers varies between 0–355 m. However, it is pertinent to mention that validation of the ice thickness results is restricted to some glaciers in the Swiss Alps region. Although, GlabTop2 model has been reported to perform very well for the glacier ice thickness estimation (Pandit and Ramsankaran, 2020).

The more accurate result can be obtained from the GPR survey; however, it was not possible for the present study due to inaccessibility. The study on ice thickness estimation using GlabTop2 and GPR investigation carried out by Kumari et al. (2021) demonstrates that, modelled thickness (average) of the Patsio glacier was estimated as 45 m whereas GPR based ice thickness measurement at four surveyed profiles (average) were found to be 62.25 m, indicating a difference of 17.25 m. The mean ice thickness values derived for the Corbassière glacier in Pennine Alps using Volta and GlabTop model are often more similar to the GPR based data as reported by Gharehchahi et al. (2020). Thus, it can be concluded that, GlabTop-2 model can provide approximate estimates of the ice thickness for wide glaciated areas that are inaccessible for the ground-based observations.

4.6 Using ensemble modelling

We employed five models, three velocity-based and two basal-shear-stress based to estimate glacier thickness using MATLAB codes. Thickness was derived through both individual and ensemble modeling, with the latter yielding more accurate results. The analysis demonstrated that ensemble modeling provided better estimates than individual approaches. The maximum thickness for L1, L2 and L3 was seen as 76.61 m, 39.79 m and 29.70 respectively (Figure 8). The maximum ice thickness was observed in the central parts (central flow lines) and near the ELA of the glacier, whereas lowest thickness (pixel value) was observed at the glacier boundaries especially at snout. Gantayat et al. (2014) have estimated the ice thickness of Gangotri glacier using GPR show that ice-thickness of the glacier varies from 15–540 m with maximum thickness in central part and minimum near the snout of the glacier was observed. The similar phenomena was observed by Ramsankaran et al. (2018) for the Chota Shigri glacier.

FIGURE 8

4.7 Volume and depth changes

Volume and depth estimation of glaciers reveal that the ice reserves contained in the glaciers has significantly reduced during the study period. The total volume of glaciers has decreased from 0.48 km³ to 0.41 km³ from base to target year. Since the volume is derived by using volume area scaling equations therefore, the pattern in changes in volume of individual glaciers is same that of area changes. Similarly, mean depth of glaciers has decreased from 57.9 m in 1980 to 55.25 m in 2020 calculated through area-based scaling technique.

4.8 Velocity estimations

Several tools and techniques were employed to derive the surface ice velocity for the Lato group of glaciers, including COSI-Corr, IMGRAFT, and GIV. Though COSI-Corr is known for better velocity estimations but in our case the results were noisy despite using all possible window size configurations available in COSI-Corr plugin. IMGRAFT, a MATLAB-based feature-tracking toolbox, though offered a good spatial distribution of velocity data but GIV produced the most realistic velocity estimates in this case with better spatial coherence and accuracy, making it the preferred approach for our analysis. Finally surface glacier velocity for the Lato group of glaciers is calculated using the GIV technique, which utilizes feature tracking from satellite or drone images. GIV ensures data accuracy by filtering velocity maps based on user-defined criteria and removing anomalies. Sentinel-2 images with 10m resolution from 2017 to 2023 were used as input. A minimum temporal baseline of approximately 11 days was selected, and image pairs with shorter time gaps were excluded from the analysis. The maximum temporal gap between image pairs extended up to 6 years, with a multi-pass frequency domain cross-correlation algorithm applied for processing. After filtering and spatial smoothing, the dataset showed a reduction in velocity outliers. During this period, the glacier-wide velocity ranged from 0.97 to 15.73 myr^-1 for L1, 1.53 to 15.63 myr^-1 for L2, and 2.09 to 14.64 myr^-1 for L3, with mean velocities of 7.67 myr^-1, 7.63 myr^-1, and 8.03 myr^-1, respectively (Figure 9).

FIGURE 9

4.9 Glacial lake dynamics

Glacial lake dynamics were analysed using multi-temporal satellite imagery for the years 1980, 1990, 2000, 2010, and 2020. The investigation reveals that the proglacial lake formed at the snout of the Lato Glacier has expanded substantially during the past four decades. The lake area increased from 0.053 ± 0.001 km² in 1980 to 0.061 ± 0.001 km² (1990), 0.068 ± 0.001 km² (2000), 0.074 ± 0.002 km² (2010), and 0.088 ± 0.002 km² in 2020, indicating an overall growth of 66% ± 5%. This corresponds to a mean rate of expansion of 0.008 ± 0.0003 km² yr^-1 over the study period (Figure 10). The average percentage change in lake area per decade was approximately 15% ± 2%, with the highest expansion (18% ± 3%) recorded between 2010 and 2020. These uncertainties reflect typical digitization and pixel-based mapping errors (∼1–2 pixels) in glacial lake extraction (Paul et al., 2013; Bolch et al., 2012). Previous studies are consistent with these findings. Kaushik et al. (2024), who analysed the hydrochemistry and area change of the Lato lake, reported a 16% increase in lake area between 1969 and 2019 and a 93% increase from 1992 to 2019, attributing the expansion to enhanced glacier melt. Similarly, the nearby Gaya Lake has undergone an extensive area expansion of 262% between 1969 and 2019 (Majeed et al., 2021). The cumulative changes in Lato glacial lake area from 1980 to 2020 are illustrated in Figure 10. The mean width of the lake was calculated as 153 ± 5 m, and the mean length as 701 ± 10 m, with uncertainties representing positional digitization error in satellite-derived boundary delineation.

FIGURE 10

4.10 Glacial lake depth, volume and peak discharge

Lake depth, volume and potential peak discharge was obtained from the equations mentioned in the Section 2.1. Equations given by Huggel et al. (2002), Fujita et al. (2013), Patel et al. (2017) was used to calculate the depth of the lake, the derived results showed a remarkable variation thus an average of all the values were taken as the mean depth of the Lato lake as suggested by Emmer et al. (2018), Chowdhury et al. (2021). The volume of the Lato lake (moraine dammed) was estimated using six equations that have been developed in various parts of the world using bathymetry datasets. The analysis of the results derived from these six equations reveals that, equations given by Evans (1986), Huggel et al. (2002), Liu et al. (2020) give results almost in a similar range. However, volume calculated through the equations of Fujita et al. (2013) shows overestimated values, on the other hand, volume obtained through (Sharma and Shukla, 2018) gives underestimated values. Thus, in this study, we have used the improved equation of volume estimation specific to moraine dammed lakes developed by Qi et al. (2022) obtained from the large lake bathymetry datasets. The volume obtained from the equation () revealed an approximate volume of the Lato lake as 1.47 × 106 m³. In order to ascertain the peak discharge of the glacial lake, three equations given in the Supplementary Table S2 were employed. Average volume 1.47 × 106 m³ calculated from the equations given in the Table was used as an input data to obtain the peak discharges from these equations individually. However, like depth and volume peak discharge values derived from the empirical equations shows a huge variation and therefore an average of all these equations was taken as a maximum peak discharge of the lake (Supplementary Table S4). The average peak discharge Q_max was calculated as 1137.22 m³. Allen et al. (2016) has suggested that equations of Fujita et al. (2013) is more reliable than the equations given by Huggel et al. (2002), Popov (1991) because in case of latter the peak discharge is overestimated. Although we did not consider the equation of Fujita et al. (2013) for the peak flow rather we used the average of all equations as suggested by Emmer et al. (2018), Chowdhury et al. (2021).

4.11 GLOF susceptibility

GLOF susceptibility of the lake located at the snout of the L1 glacier in Lato basin was assessed using multi criteria decision-based method. Parameters/factors were selected from the various potential factors that govern the GLOF mechanism. The parameters mentioned in the Table represents these factors for example f1, f2, f3, and f6 represent the glacier and lake characteristics, f4 and f7 represents the dam condition and f5 and represents the surrounding characteristics (Table 2). The area of the lake was estimated as 0.088 km² as per the recent estimates (2020), which is greater than the threshold area suggested by Khadka et al. (2022) to assess the GLOF susceptibility. The lake is classified as moraine dammed lake, located at the snout of the L1 glacier. Lakes dammed by moraines are considered as potentially more dangerous as compared to the other types of lake because weak moraines are subjected to sudden breaching in case lake is hit by any GLOF triggering factors (Snow/ice avalanche, snout calving, Rock fall, cloudburst). In this study, we considered lake area to determine the magnitude of GLOF rather than lake volume because all the existing empirical equations for volume estimation are area dependent (Fujita et al., 2013; Khadka et al., 2022). The slope of the dam front (distal slope) similar to average lake front area slope and it is generally considered as relevant to moraine-dammed lakes representing the possibilities of the erosion, incision and failure. The dam front slope for the selected lake was estimated to be greater than 10°, which shows the high hazard level of the parameters with an index value of 1. The area of the lake has revealed significant growth (56%) over the last 40 years which is greater than the threshold of high hazard value given by Che et al. (2014), Jain et al. (2012). The lake expansion has caused due to the retreat of the feeding glacier snout, which provides a continuous meltwater supply to the lake. With the increasing temperatures and melting of glaciers such lakes may expand faster and therefore increasing the threat of GLOF. The lake is also susceptible to the potential ice and rock avalanches due its steep slope of the feeding glacier and side valleys.

The GLOF susceptibility index shows high level of hazard with high index value for 5 factors such as f2, f3, f4, f5 and f7, whereas medium hazard level was found for the f1 and f6 factors. The lake moraine dammed, located at the snout of the glacier with an area greater than 0.45 km² and expansion rate more than 50%. Overall, the lake has a susceptibility index value of 0.82 which indicates that lake is highly susceptible to GLOF and therefore it should be taken as a case study to further evaluate the possible GLOF impacts on the downstream areas using the combined approaches of remote sensing GIS and hydrodynamic modeling.

4.12 GLOFs simulation using HEC-RAS

In this study, we used HEC-RAS two-dimensional unsteady flow analysis to simulate the potential breach of Lato lake in Ladakh. The simulation modeled the outburst flow over 23 km, extending from the lake to Miru village. The GLOF simulations for Scenario 1 and Scenario 2 indicate significant variations in peak discharge (Figure 11). In scenario 1, the 100% lake volume is assumed to breach, leading to a catastrophic release of water. For this worst-case scenario, the breach width and breach formation time was calculated to be 43.56 m and 0.13 h respectively (Table 5). The resulting hydrograph indicates a peak discharge of 3726 m³s^-1, occurring rapidly within 7.8 min of the event (Figure 11). This high discharge rate suggests severe downstream impacts, with substantial flooding, erosion, and potential infrastructure damage in the nearby villages. In scenario 2, 75% of the lake volume is considered to breach, simulating a partial failure. The peak discharge for this scenario is 2309 m³s^-1, significantly lower than Scenario 1. Although this scenario results in a lower flood magnitude, it still poses substantial risks to downstream communities and infrastructure. The results highlight that Scenario 1 revealed a more extreme event, leading to higher discharge, deeper inundation, and greater velocity impacts (Figure 12). Scenario 2, while still hazardous, results in a comparatively milder flood wave. This difference in peak discharges highlights the importance of breach volume in determining the flood severity.

FIGURE 11

FIGURE 12

The inundation maps shows that the flood wave propagates rapidly downstream, affecting settlements and infrastructure along the valley (Figure 12). Velocity maps indicate maximum flow velocities exceeding 18 ms^-1, posing a significant threat to downstream communities. The comparison between the two scenarios highlights the sensitivity of flood magnitude to breach parameters, with Scenario 1 leading to a more destructive flood due to maximum lake drainage. In Scenario 1 (Figure 12a) and (Figure 12c), the peak flood depth in the Lato village reaches approximately 3.5 m within the initial 45 min, followed by a gradual recession. In Scenario 2 (Figure 12b) and (Figure 12c), a slightly lower peak depth of around 2.5 m is observed due to differences in breach parameters. In Miru village, located further downstream, the flood depth peaks at 2 m in Scenario 1 within 1 h and 55 min, whereas in Scenario 2, it reaches approximately 1.1 m after 3 h and 25 min (Figure 12a), (Figure 12b) and (Figure 12c). Similarly, the peak flow velocity in Lato village for Scenario 1 is 5.7 ms^-1 (Figure 12d) and (Figure 12f), whereas Scenario 2 results in a slightly reduced velocity of 3.8 ms^-1 (Figure 12e) and (Figure 12f). As the flood wave travels downstream to Miru village, the velocity reduces significantly, reaching 2.6 ms^-1 and 1.6 ms^-1 for Scenarios 1 and 2, respectively (Figure 12d), (Figure 12e) and (Figure 12f). The results highlight that Lato faces the most immediate and severe impacts due to its proximity to the lake, experiencing high flood depths and velocities. In contrast, Miru experiences an attenuated flood wave with reduced intensity.

4.13 Sensitivity analysis

To evaluate the influence of uncertainty in breach dynamics, a sensitivity analysis was performed by varying breach width and failure time. Six different cases were considered to assess how these breach parameters influence the breach hydrograph. Variations in time of failure (t_f) (30 min, 45 min, and 60 min) were analyzed keeping the constant breach width (b_w) of 100 m. For constant b_w, a decrease in t_f from 60 to 30 min led to an increase in peak discharge from ∼1000 m³s^-1–∼1800 m³s^-1 (Figure 13a). Therefore, a shorter failure time resulted in a more abrupt and intense flood wave, while a longer failure time led to a more gradual rise in discharge. Simulations were run for different breach widths of (50 m, 75 m, and 100 m) with a constant time of failure of 30 min. The highest discharge (∼1700 m³s^-1) was observed for a b_w of 100 m and t_f of 30 min, while the lowest (∼1150 m³s^-1) occurred for a b_w of 50 m (Figure 13b). A wider breach leds to an increase in peak discharge and a faster flood wave propagation downstream, which results increases in flood depth and velocity at both Lato and Miru villages. From this assessment, it is evident that GLOF hydrographs are more sensitive to breach formation time compared to breach width. A shorter breach formation time significantly amplifies the peak discharge, whereas variations in breach width have a comparatively moderate influence.

FIGURE 13

4.14 GLOF risk and exposure assessment

GLOFs pose a significant risk to downstream communities, infrastructure, and ecosystems in high-mountain regions. The risk is driven by the sudden release of large volumes of water from glacial lakes, often triggered by events such as ice or rock avalanches, moraine dam failure, or extreme weather conditions. Key factors contributing to GLOF risk include the size and stability of the glacial lake, the volume of stored water, the topography of the surrounding area, and the proximity of vulnerable elements such as settlements, roads, agricultural land, and hydropower infrastructure. In the context of the study area, the exposure assessment highlights the potential impacts of GLOFs under different scenarios. For instance, Scenario 1 indicates a higher risk, with a larger inundated area (2.3 km²) affecting critical infrastructure such as 2.2 km of roads, 13 buildings, and 2.1 ha of agricultural land, as well as endangering a population of 58 individuals. In Scenario 2, the risk is comparatively lower, with an inundated area of 1.6 km², impacting 1.7 km of roads, 8 buildings, 1.5 ha of agricultural land, and 33 people. Mitigating GLOF risk requires a multi-faceted approach, including continuous monitoring of glacial lakes, early warning systems, community preparedness programs, and structural measures such as controlled drainage or reinforcement of unstable moraine dams. Understanding the specific vulnerabilities and exposure of elements at risk is crucial for developing effective risk reduction strategies and ensuring the safety and resilience of downstream communities.

4.15 Uncertainty analysis

4.15.1 Glacier surface velocity

The uncertainty linked to glacier surface velocity, derived from feature tracking of satellite imagery, stems from multiple factors including errors in image co-registration, image distortions, false matches, and imprecise subpixel displacement estimations. This study delves into the uncertainty linked with glacier surface velocity through testing on the stable terrain of the Lato group of Glaciers. The established velocity of stable terrain, determined to be 0 m yr⁻¹ serves as a reference point, enabling the identification of any deviations from this value due to data errors (Singh et al., 2023; Gantayat et al., 2014). Subsequently, five polygons representing ‘stable ground’ were identified, characterized by the absence of ice and snow as well as the absence of landsliding. The mean of the velocity raster was 0.28 whereas the max velocity was observed 3.29 m yr^-1. The standard deviation of 0.51 was observed in velocity across these stable regions, providing an estimate of the uncertainty in velocity associated with the GIV feature tracking method.

4.15.2 Glacier ice thickness

The main contributors to uncertainty in thickness include inaccuracies in surface velocity, variability in the shape factor, fluctuations in glacier ice density, and errors in slope calculations. The collective relative uncertainty in estimating glacier ice volume is determined using an equation derived from taking the natural logarithm of both sides of Equation 1 and then differentiating

Surface velocity uncertainty arises from orthorectification and georectification errors, yielding a cumulative error of 3–9 m. Lacking Ground velocity 3.5 m per year is assumed from prior research. The assumed range for the shape factor (f) leads to an uncertainty of 0.2. The creep parameter (A) is set at 3.24 × 10–24 Pa ^-3 s^-1, and a uniform ice density (ρ) of 900 kg/m-3, with a potential decrease to 850 kgm^-3, resulting in a 5.5% uncertainty when d is 50 kg/m^-3(Sattar et al., 2019b). Moreover, Copernicus 30m data was employed for elevation and slope analysis, wherein the inherent uncertainty errors, such as RMSE and MAE ranged from 6% to 44% (Ghannadi et al., 2023). In addition to this an overall thickness uncertainty estimation was carried out similar to velocity estimation, for some (same five) ice free regions, the thickness of was calculated which ideally must be zero but shows a mean thickness of 3.02 m whereas the max ice thickness reached up to 8.03m with a standard deviation of 2.83 for the whole ice-free raster. This provides an estimate of the uncertainty in the thickness modelling approach used in the present study.

This study offers a rare high-resolution, process-integrated view of accelerating cryospheric change and emerging GLOF risk in the Trans-Himalayan region, centered on the Lato basin in Ladakh. The observed glacier retreat, lake expansion, and mass loss are consistent with regional trends reported for the Himalaya and HMA, yet the rate and spatial manifestation of these changes in Lato are particularly concerning given its steep topography, arid climate, and downstream vulnerability. The ∼11.5% glacier area loss and substantial snout retreat between 1980 and 2020—paired with a negative geodetic mass balance of −0.33 m w.e. yr^-1—underscore the basin’s sensitivity to warming, even in regions receiving limited monsoonal precipitation. The integration of remote sensing, velocity tracking, ensemble ice-thickness modeling, and hydrodynamic simulation represents a novel, comprehensive approach in Himalayan glacier–lake systems. GIV provided spatially coherent velocity fields, with mean velocities of 7–8 m yr^-1, and ensemble modeling approaches produced mean ice thickness estimates ranging between 58 and 76 m. These data not only improve the accuracy of water storage estimates but also inform downstream hazard modeling. Our results highlight that while GlabTop2 yielded reliable mean thickness values, ensemble modeling captured more variability and nuance across different glacier geometries—indicating the importance of multi-model convergence for critical glacier parameters.

The rapid growth of the Lato proglacial lake—by 66% over four decades—follows the pattern observed in adjacent basins such as Gaya and Chandra, but with one critical difference: the current susceptibility index of 0.82 indicates a notably higher hazard potential. This is driven by multiple converging factors: continuous snout retreat providing sustained meltwater input, steep dam-front slopes, potential avalanche-prone topography, and the unconsolidated moraine dam structure. The susceptibility ranking aligns with thresholds proposed in previous GLOF vulnerability frameworks, yet highlights the need to re-calibrate them for smaller lakes that evolve in highly dynamic glacier systems. Simulated GLOF scenarios through HEC-RAS modeling reveal critical exposure hotspots, notably in the Lato and Miru settlements, with predicted peak discharges exceeding 3700 m³ s^-1 in full breach conditions. These values fall within the range of documented historical GLOFs in the Himalaya, reinforcing the notion that even modest-sized lakes can generate catastrophic downstream impacts when coupled with steep, confined valleys. The sensitivity analysis emphasizes that breach formation time—more than breach width—is a dominant control on peak discharge, calling attention to the importance of dynamic triggers such as ice calving, seismic activity, or extreme precipitation events.

Despite the methodological strengths, several uncertainties merit discussion. Ice thickness modeling, though enhanced via ensemble approaches, remains constrained by the absence of ground-penetrating radar (GPR) validation, and satellite-derived glacier velocities can be affected by surface conditions, co-registration errors, or temporal baselines. Similarly, empirical equations for lake volume and peak discharge estimation—though widely used—may not fully capture local geometric complexity. Nonetheless, the converging evidence across methods and datasets affirms the directionality and severity of the risk. The implications of this study extend beyond the Lato basin. As warming accelerates, similar moraine-dammed lakes are forming and expanding across HMA, particularly in debris-covered glacier zones where field access is limited and response capacity is low. The Lato case exemplifies how integrated remote and modeled data can fill critical knowledge gaps in these frontier systems. More importantly, it demonstrates that small, high-altitude lakes—often overlooked in broad-scale inventories—can exhibit disproportionately high hazard levels.

Given the high GLOF susceptibility of the Lato glacial lake and the observed rapid glacier dynamics, we strongly recommend establishing a continuous monitoring system. Early warning systems should be prioritized for downstream villages such as Lato and Miru. Field-based validation of ice thickness and lake depth should be undertaken to refine hazard models. Future work should prioritize high-resolution bathymetric surveys, GPR-based thickness validation, and long-term in situ monitoring of glacier–lake dynamics, particularly in catchments with downstream settlements and infrastructure. Moreover, cross-disciplinary approaches combining glaciology, hydrodynamics, hazard science, and local knowledge systems are essential to inform targeted adaptation strategies, from early warning systems to infrastructure zoning. As mountain regions face compound climate risks, scientific assessments like this must underpin urgent action on climate resilience and disaster risk reduction.

5 Conclusion

This study provides a comprehensive, multi-parameter assessment of glacier dynamics and emerging GLOF risk in the Lato basin of Ladakh, a critically understudied region in the Trans-Himalaya. Using remote sensing data spanning 1980–2023, we observed ∼11.5% glacier area loss, significant terminus retreat (∼171 m), and a decrease in glacier volume from 0.48 km³ to 0.41 km³. Glacier surface velocities, derived through GIV and other feature-tracking methods, indicated a slowing trend, with mean values ranging from 7.63 to 8.03 m yr^-1 between 2017 and 2023. Ice thickness estimations using both GlabTop2 and ensemble models revealed thinning glaciers, with average thicknesses of 76 m and 58 m, respectively. Geodetic mass balance calculations pointed to a persistent negative mass balance trend of −0.33 m w.e. yr^-1 since the early 2000s. The associated proglacial lake has expanded by approximately 66% since 1980, with a current surface area of 0.088 km² and a robustly estimated volume of 1.47 × 10⁶ m³. A GLOF susceptibility index of 0.82—calculated using AHP-based multi-criteria analysis—classifies the Lato lake as highly susceptible to outburst flooding. Hydrodynamic simulations using HEC-RAS under full and partial breach scenarios forecast potentially destructive peak discharges of 3726 m³s^-1 and 2309 m³s^-1, posing serious threats to downstream settlements over a 23 km stretch. These findings highlight the pressing need for integrating scientific understandings into local and regional policy frameworks. Region-specific strategies such as real-time glacier and lake monitoring, installation of ESW, community-based disaster preparedness, and restrictions on infrastructure development in vulnerable zones are essential to reduce the potential impact of future GLOFs. Moreover, this integrative methodology—combining remote sensing, glacier modeling, hazard indexing, and hydraulic simulations—can serve as a replicable framework for other high-risk glacial lake basins across the Himalaya.

References

• 1

  AbowardaA. S.BaiL.ZhangC.LongDiLiX.HuangQiet al (2021). Generating surface soil moisture at 30 m spatial resolution using both data fusion and machine learning toward better water resources management at the field scale. Remote Sens. Environ.255, 112301. 10.1016/j.rse.2021.112301

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AggarwalS.RaiS. C.ThakurP. K.EmmerA. (2017). Inventory and recently increasing GLOF susceptibility of glacial lakes in Sikkim, Eastern Himalaya. Geomorphology295, 39–54. 10.1016/j.geomorph.2017.06.014

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AhmadS. T.AhmedR.WaniG. F.SharmaP.AhmedP. (2022). Glacier changes in Sind basin (1990–2018) of North-western Himalayas using earth observation data. Model. Earth Syst. Environ.8 (2), 2567–2579. 10.1007/s40808-021-01246-w

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  AllenS. K.LinsbauerA.RandhawaS. S.HuggelC.RanaP.KumariA. (2016). Glacial Lake outburst flood risk in Himachal Pradesh, India. An integrative and anticipatory approach considering current and future threats. Nat. Hazards84 (3), 1741–1763. 10.1007/s11069-016-2511-x

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  AnaconaP. I.MackintoshA.NortonK. (2015). Reconstruction of a glacial lake outburst flood (GLOF) in the Engaño Valley, Chilean Patagonia. Lessons for GLOF risk management. Sci. Total Environment527-528, 1–11. 10.1016/j.scitotenv.2015.04.096

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  AwasthiS.VaradeD.Kumar ThakurP.KumarA.SinghH.JainK.et al (2022). Development of a novel approach for snow wetness estimation using hybrid polarimetric RISAT-1 SAR datasets in North-Western Himalayan region. J. Hydrology612 (8), 128252. 10.1016/j.jhydrol.2022.128252

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  AzamM. F.WagnonP.BerthierE.VincentC.FujitaK.KargelJ. S. (2018). Review of the status and mass changes of Himalayan-Karakoram glaciers. J. Glaciol.64 (243), 61–74. 10.1017/jog.2017.86

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  BahrD. B.PfefferW. T.KaserG. (2015). A review of volume‐area scaling of glaciers. Rev. Geophys.53 (1), 95–140. 10.1002/2014RG000470

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  BajracharyaS. R.MaharjanS. B.ShresthaF. (2014). The status and decadal change of glaciers in Bhutan from the 1980s to 2010 based on satellite data. Ann. Glaciol.55 (66), 159–166. 10.3189/2014aog66a125

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  BarryR. G. (2006). The status of research on glaciers and global glacier recession. A review. Prog. Phys. Geogr. Earth Environ.30 (3), 285–306. 10.1191/0309133306pp478ra

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  BerthierE.ArnaudY.VincentC.RemyF. (2006). Biases of SRTM in high-mountain areas: implications for the monitoring of glacier volume changes. Geophys. Res. Lett.33, L08502. 10.1029/2006GL025862

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  BhambriR.MisraA.KumarA.GuptaA. K.VermaA.TiwariS. K. (2018). Glacier lake inventory of Himachal Pradesh. Himal. Geol.39 (1), 1–32.

  • Google Scholar
  • View reference in article

• 13

  BhardwajA.AliS. N.SamL.PandeyP. (2025). Time to diffuse the “ticking time bombs” of the Himalaya. Phys. Chem. Earth, Parts A/B/C139, 103908. 10.1016/j.pce.2025.103908

  • CrossRef
  • Google Scholar

• 14

  BhattacharjeeS.PandeyA. C.GargR. D. (2023). Long-term estimation of glacier mass balance using geospatial techniques in Western Himalayas, Ladakh, India. Quat. Sci. Adv.12, 100118. 10.1016/j.qsa.2023.100118

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  BhattacharyaA.BolchT.MukherjeeK.KingO.MenounosB.KapitsaV.et al (2021). High Mountain Asian glacier response to climate revealed by multi-temporal satellite observations since the 1960s. Nat. Communications12 (1), 4133. 10.1038/s41467-021-24180-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  BolchT.KulkarniA.KääbA.HuggelC.PaulF.CogleyJ. G.et al (2012). The state and fate of Himalayan glaciers. Science336 (6079), 310–314. 10.1126/science.1215828

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  BrunF.BerthierE.WagnonP.KääbA.TreichlerD. (2017). A spatially resolved estimate of high Mountain Asia glacier mass balances, 2000-2016. Nat. Geosci.10 (9), 668–673. 10.1038/ngeo2999

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  CampbellJ. G. (2005). Inventory of glaciers, glacial lakes and the identification of potential glacial lake outburst floods (GLOFs) affected by global warming in the mountains of India, Pakistan and China/Tibet Autonomous Region: final report for APN project 2004-03-CMY-Campbell. International Centre for Integrated Mountain Development.

  • Google Scholar
  • View reference in article

• 19

  CarrivickJ. L.TweedF. S. (2013). Proglacial lakes. Character, behaviour and geological importance. Quat. Sci. Rev.78 (1), 34–52. 10.1016/j.quascirev.2013.07.028

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  CarrivickJ. L.TweedF. S. (2016). A global assessment of the societal impacts of glacier outburst floods. Glob. Planet. Change144 (2), 1–16. 10.1016/j.gloplacha.2016.07.001

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  CheT.TaoX.LinX.LiouY.-A. (2014). Changes in glaciers and glacial lakes and the identification of dangerous glacial lakes in the Pumqu River Basin, Xizang (Tibet). Adv. Meteorology2014. 10.1155/2014/903709

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  ChenM.ChenY.FangG.ZhengG.LiZ.LiY.et al (2024). Risk assessment of glacial lake outburst flood in the Central Asian Tienshan Mountains. Npj Clim. Atmos. Sci.7 (1), 907. 10.1038/s41612-024-00755-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  ChevuturiA.DimriA. P.ThayyenR. J. (2018). Climate change over Leh (Ladakh), India. Theor. Appl. Climatol.131, 531–545. 10.1007/s00704-016-1989-1

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  ChowdhuryA.KroczekT.DeS. K.VilímekV.SharmaM. C.DebnathM. (2021). Glacial lake evolution (1962–2018) and outburst susceptibility of gurudongmar lake complex in the Tista basin, Sikkim Himalaya (India). Water13 (24), 3565. 10.3390/w13243565

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  CogleyJ. G. (2016). Glacier shrinkage across High Mountain Asia. Ann. Glaciol.57 (71), 41–49. 10.3189/2016AoG71A040

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  CoppinP.JonckheereI.NackaertsK.MuysB.LambinE. (2004). Review ArticleDigital change detection methods in ecosystem monitoring: a review. Int. Journal Remote Sensing25 (9), 1565–1596. 10.1080/0143116031000101675

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  CuffeyK. M.PatersonW. S. B. (2010). The physics of glaciers. 4th edn.Oxford: Butterworth-Heinemann.

  • Google Scholar
  • View reference in article

• 28

  DeBEERC. M.SharpM. J. (2009). Topographic influences on recent changes of very small glaciers in the Monashee Mountains, British Columbia, Canada. J. Glaciol.55 (192), 691–700. 10.3189/002214309789470851

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  DehecqA.GourmelenN.GardnerA. S.BrunF.GoldbergD.NienowP. W.et al (2019). Twenty-first century glacier slowdown driven by mass loss in high Mountain Asia. Nat. Geosci.12 (1), 22–27. 10.1038/s41561-018-0271-9

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  DobhalD. P.GerganJ. T.ThayyenR. J. (2004). Recession and morphogeometrical changes of Dokriani glacier (1962-1995) Garhwal Himalaya, India. Curr. SCIENCE-BANGALORE-86 (5), 692–696.

  • Google Scholar
  • View reference in article

• 31

  DubeyJ.SinghD. S.SatiS. P. (2025). “Cryosphere response to climate change and disaster risk reduction in the arid Northwestern Ladakh himalaya,” in Ecosystem-Based approaches for resilience building in Himalayan landscapes: insights for climate change and disaster risk reduction (Singapore: Springer Nature Singapore), 3–27.

  • Google Scholar
  • View reference in article

• 32

  EmmerA.VilímekV.ZapataM. L. (2018). Hazard mitigation of glacial lake outburst floods in the Cordillera Blanca (Peru): the effectiveness of remedial works. J. Flood Risk Manag.11, S489–S501. 10.1111/jfr3.12241

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  EmmerA.HarrisonS.MergiliM.AllenS.FreyH.HuggelC. (2020). 70 years of lake evolution and glacial lake outburst floods in the Cordillera Blanca (Peru) and implications for the future. Geomorphology365 (5), 107178. 10.1016/j.geomorph.2020.107178

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  EvansS. G. (1986). The maximum discharge of outburst floods caused by the breaching of man-made and natural dams. Can. Geotechnical J.23 (3), 385–387. 10.1139/t86-053

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  FarinottiD.HussM.BauderA.FunkM.TrufferM. (2009). A method to estimate the ice volume and ice-thickness distribution of alpine glaciers. J. Glaciol.55 (191), 422–430. 10.3189/002214309788816759

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  FarinottiD.HussM.FürstJ. J.LandmannJ.MachguthH.MaussionF.et al (2019). A consensus estimate for the ice thickness distribution of all glaciers on Earth. Nat. Geosci.12 (3), 168–173. 10.1038/s41561-019-0300-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  FreyH.HaeberliW.LinsbauerA.HuggelC.PaulF. (2010). A multi-level strategy for anticipating future glacier lake formation and associated hazard potentials. Nat. Hazards Earth Syst. Sci.10 (2), 339–352. 10.5194/nhess-10-339-2010

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  FreyH.MachguthH.HussM.HuggelC.BajracharyaS.BolchT.et al (2014). Estimating the volume of glaciers in the Himalayan–Karakoram region using different methods. Cryosphere8 (6), 2313–2333. 10.5194/tc-8-2313-2014

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  FroehlichD. C. (1995). Peak outflow from Breached embankment dam. J. Water resour. Plan. Manag.121, 90–97. 10.1061/(asce)0733-9496(1995)121:1(90)

  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  FujitaK.SakaiA.TakenakaS.NuimuraT.SurazakovA. B.SawagakiT.et al (2013). Potential flood volume of Himalayan glacial lakes. Nat. Hazards Earth Syst. Sci.13 (7), 1827–1839. 10.5194/nhess-13-1827-2013

  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  FurianW.MaussionF.SchneiderC. (2022). Projected 21st-Century glacial Lake evolution in high Mountain Asia. Front. Earth Sci.10, 125. 10.3389/feart.2022.821798

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  GantayatP.KulkarniA. V.SrinivasanJ. (2014). Estimation of ice thickness using surface velocities and slope: case study at Gangotri Glacier, India. J. Glaciol.60 (220), 277–282. 10.3189/2014jog13j078

  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  GhannadiM. A.AlebooyeS.IzadiM.GhanadiA. (2023). Vertical accuracy assessment of Copernicus DEM (case study: tehran and Jam cities). ISPRS annals of the photogrammetry. Remote Sens. Spatial Inf. Sci.10, 209–214.

  • Google Scholar
  • View reference in article

• 44

  GharehchahiS.JamesW. H.BhardwajA.JensenJ. L.SamL.BallingerT. J.et al (2020). Glacier ice thickness estimation and future Lake formation in Swiss Southwestern alps—the upper Rhône catchment: a VOLTA application. Remote Sens.12 (20), 3443. 10.3390/rs12203443

  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  GhoshT. (2024). Low snow levels in the Himalaya are putting lives at risk.

  • Google Scholar

• 46

  GopikaJ. S.KulkarniA. V.PrasadV.SrinivasaluP.RamanA. (2021). Estimation of glacier stored water in the Bhaga basin using laminar flow and volume-area scaling methods. Remote Sens. Appl. Soc. Environ.24, 100656. 10.1016/j.rsase.2021.100656

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  HaeberliW.HoelzleM. (1995). Application of inventory data for estimating characteristics of and regional climate-change effects on mountain glaciers: a pilot study with the European alps. Ann. Glaciology21, 206–212. 10.3189/s0260305500015834

  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  HeR.ShangguanD.ZhaoQ.ZhangS.JinZ.QinY.et al (2024). Effective identification of debris-covered glaciers in Western China using multiple machine-learning algorithms. Sci. Total Environ.955, 176946. 10.1016/j.scitotenv.2024.176946

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  HeidT.KääbA. (2012). Evaluation of existing image matching methods for deriving glacier surface displacements globally from optical satellite imagery. Remote Sens. Environ.118, 339–355. 10.1016/j.rse.2011.11.024

  • CrossRef
  • Google Scholar
  • View reference in article

• 50

  HockR.HussM. (2021). “Glaciers and climate change,” in Climate change (Elsevier), 157–176.

  • Google Scholar
  • View reference in article

• 51

  HockR.RasulG.AdlerC.CáceresB.GruberS.HirabayashiY.et al (2019). “High mountain areas,” in IPCC special report on the ocean and cryosphere in a changing climate. Editors PörtnerH-O.RobertsD. C.Masson-DelmotteV.ZhaiP.TignorM.PoloczanskaE., 131–202.

  • Google Scholar

• 52

  HuggelC.KaabA.HaeberliW.TeysseireP.PaulF. (2002). Remote sensing-based assessment of hazards from glacier lake outbursts: a case study in the Swiss Alps. Can. Geotech. J.39, 316–330. 10.1139/t01-099

  • CrossRef
  • Google Scholar
  • View reference in article

• 53

  HuggelC.KääbA.HaeberliW.KrummenacherB. (2003). Regional-scale GIS-models for assessment of hazards from glacier Lake outbursts: evaluation and application in the Swiss alps. Nat. Hazards Earth Syst. Sci.3 (6), 647–662. 10.5194/nhess-3-647-2003

  • CrossRef
  • Google Scholar
  • View reference in article

• 54

  HuggelC.CareyM.EmmerA.FreyH.Walker-CrawfordN.Wallimann-HelmerI. (2020). Anthropogenic climate change and glacier lake outburst flood risk: local and global drivers and responsibilities for the case of lake Palcacocha, Peru. Nat. Hazards Earth Syst. Sci.20, 2175–2193. 10.5194/nhess-20-2175-2020

  • CrossRef
  • Google Scholar
  • View reference in article

• 55

  HugonnetR.McNabbR.BerthierE.MenounosB.NuthC.GirodL.et al (2021). Accelerated global glacier mass loss in the early twenty-first century. Nature592 (7856), 726–731. 10.1038/s41586-021-03436-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 56

  HussM. (2013). Density assumptions for converting geodetic glacier volume change to mass change. Cryosphere7 (3), 877–887. 10.5194/tc-7-877-2013

  • CrossRef
  • Google Scholar
  • View reference in article

• 57

  HussM.HockR. (2015). A new model for global glacier change and sea-level rise. Front. Earth Sci.3, 54. 10.3389/feart.2015.00054

  • CrossRef
  • Google Scholar
  • View reference in article

• 58

  ICIMOD (2011). Glacial Lakes and Glacial Lake Outburst Floods in Nepal. Kathmandu: ICIMOD.

  • Google Scholar

• 59

  ImmerzeelW. W.Van BeekL. P. H.KonzM.ShresthaA. B.BierkensM. F. P. (2012). Hydrological response to climate change in a glacierized catchment in the Himalayas. Clim. Change110, 721–736. 10.1007/s10584-011-0143-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 60

  ImmerzeelW. W.LutzA. F.AndradeM.BahlA.BiemansH.BolchT.et al (2020). Importance and vulnerability of the world's water towers. Nature577 (7790), 364–369. 10.1038/s41586-019-1822-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 61

  IslamN.PatelP. P. (2022). Inventory and GLOF hazard assessment of glacial lakes in the Sikkim Himalayas, India. Geocarto Int.37 (13), 3840–3876. 10.1080/10106049.2020.1869332

  • CrossRef
  • Google Scholar
  • View reference in article

• 62

  JainS. K.LohaniA. K.SinghR. D.ChaudharyA.ThakuralL. N. (2012). Glacial lakes and glacial lake outburst flood in a Himalayan basin using remote sensing and GIS. Nat. Hazards62, 887–899. 10.1007/s11069-012-0120-x

  • CrossRef
  • Google Scholar
  • View reference in article

• 63

  JawakS. D.KumarS.LuisA. J.BartanwalaM.TummalaS.PandeyA. C. (2018). “Evaluation of geospatial tools for generating accurate glacier velocity maps from optical remote sensing data,”, 2. MDPI, 341. 10.3390/ecrs-2-05154

  • CrossRef
  • Google Scholar
  • View reference in article

• 64

  KarimiN.FarokhniaA.KarimiL.EftekhariM.GhalkhaniH. (2012). Combining optical and thermal remote sensing data for mapping debris-covered glaciers (Alamkouh Glaciers, Iran). Cold Regions Sci. Technol.71, 73–83. 10.1016/j.coldregions.2011.10.004

  • CrossRef
  • Google Scholar
  • View reference in article

• 65

  KaushikS.RafiqM.DharpureJ. K.HowatI.MoortgatJ.JoshiP. K.et al (2024). Increasing risk of glacial lake outburst flood in Sikkim, Eastern Himalaya under climate warming. Remote Sens. Appl. Soc. Environ.36 (6), 101286. 10.1016/j.rsase.2024.101286

  • CrossRef
  • Google Scholar
  • View reference in article

• 66

  KennedyR. E.AndréfouëtS.CohenW. B.GómezC.GriffithsP.HaisM.et al (2014). Bringing an ecological view of change to andsat‐based remote sensing. Front. Ecol. Environ.12 (6), 339–346. 10.1890/130066

  • CrossRef
  • Google Scholar
  • View reference in article

• 67

  KhadkaN.GhimireS. K.SharmaS.HamalK. (2022). Rapidly expanding glacial lakes in Nepal Himalaya. Jalawaayu2 (1), 45–55. 10.3126/jalawaayu.v2i1.45393

  • CrossRef
  • Google Scholar
  • View reference in article

• 68

  KhanA. (2021). Cloud shadow detection removal for satellite supportive health care systems: research solution towards Australian Bushfire. J. Ambient Intell. Humaniz. Comput.12 (11), 10239–10251. 10.1007/s12652-020-02793-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 69

  KraaijenbrinkP. D. A.BierkensM. F. P.LutzA. F.ImmerzeelW. W. (2017). Impact of a global temperature rise of 1.5 degrees Celsius on Asia's glaciers. Nature549 (7671), 257–260. 10.1038/nature23878

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 70

  KumariS.PanditA.RamsankaranR. A. A. J.SohebM.AngchukT.RamanathanA. L. (2021). Modelling ice thickness distribution and volume of Patsio Glacier in Western Himalayas. J. Earth Syst. Sci.130, 1–14. 10.1007/s12040-021-01643-w

  • CrossRef
  • Google Scholar
  • View reference in article

• 71

  LachaiseM.BachmannM.SchweisshelmB.FritzT. (2021). “The Tandem-X change DEM: status of the change raw DEMs production,” in Proceedings of the 2021 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 8022–8025.

  • Google Scholar
  • View reference in article

• 72

  LanghammerL.GrabM.BauderA.MaurerH. (2019). Glacier thickness estimations of alpine glaciers using data and modeling constraints. Cryosphere13 (8), 2189–2202. 10.5194/tc-13-2189-2019

  • CrossRef
  • Google Scholar
  • View reference in article

• 73

  LiJ.LiZ.-W.HuJ.WuL. X.LiX.GuoL.et al (2021). Investigating the bias of TanDEM-X digital elevation models of glaciers on the Tibetan Plateau: impacting factors and potential effects on geodetic mass-balance measurements. J. Glaciol.67 (264), 613–626. 10.1017/jog.2021.15

  • CrossRef
  • Google Scholar
  • View reference in article

• 74

  LinsbauerA.PaulF.HaeberliW. (2012). Modeling glacier thickness distribution and bed topography over entire mountain ranges with GlabTop: application of a fast and robust approach. J. Geophys. Res. Earth Surf.117 (F3), F03007. 10.1029/2011JF002313

  • CrossRef
  • Google Scholar
  • View reference in article

• 75

  LinsbauerA.FreyH.HaeberliW.MachguthH.AzamM. F.AllenS. (2016). Modelling glacier-bed overdeepenings and possible future lakes for the glaciers in the Himalaya—Karakoram region. Ann. Glaciol.57 (71), 119–130. 10.3189/2016AoG71A627

  • CrossRef
  • Google Scholar
  • View reference in article

• 76

  LiuQ.MayerC.WangX.NieY.WuK.WeiJ.et al (2020). Interannual flow dynamics driven by frontal retreat of a lake-terminatin.

  • Google Scholar
  • View reference in article

• 77

  MaanyaU. S.KulkarniA. V.TiwariA.BharE. D.SrinivasanJ. (2016). Identification of potential glacial lake sites and mapping maximum extent of existing glacier lakes in Drang Drung and Samudra Tapu glaciers, Indian Himalaya. Curr. Sci.111, 553–560. 10.18520/cs/v111/i3/553-560

  • CrossRef
  • Google Scholar
  • View reference in article

• 78

  MajeedU.RashidI.NajarN. A.GulN. (2021). Spatiotemporal dynamics and geodetic mass changes of glaciers with varying debris cover in the Pangong Region of trans-Himalayan Ladakh, India between 1990 and 2019. Front. Earth Sci.9, 748107. 10.3389/feart.2021.748107

  • CrossRef
  • Google Scholar
  • View reference in article

• 79

  MankotiaS.AhmedR.SiddiquiM. A.AhmadI.RamizM.NasrinT. (2025). Glacier retreat and morphological changes in the Suru Sub Basin of Ladakh region from 1992 to 2023. Front. Eng. Built Environ.5 (2), 152–172. 10.1108/febe-12-2024-0087

  • CrossRef
  • Google Scholar
  • View reference in article

• 80

  McFeetersS. K. (2013). Using the normalized difference water index (NDWI) within a geographic information system to detect swimming pools for mosquito abatement: a practical approach. Remote Sens.5 (7), 3544–3561. 10.3390/rs5073544

  • CrossRef
  • Google Scholar

• 81

  MeierM. F. (1962). Proposed definitions for glacier mass budget terms. J. Glaciol.4 (33), 252–263. 10.3189/s0022143000027544

  • CrossRef
  • Google Scholar

• 82

  MirR. A.JainS. K.LohaniA. K.SarafA. K. (2018). Glacier recession and glacial lake outburst flood studies in Zanskar basin, western Himalaya. J. Hydrology564, 376–396. 10.1016/j.jhydrol.2018.05.031

  • CrossRef
  • Google Scholar

• 83

  MishraP. K.ThayyenR. J.SinghH.DasS.NemaM. K.KumarP. (2022). Assessment of cloudbursts, extreme rainfall and vulnerable regions in the Upper Ganga basin, Uttarakhand, India. Int. J. Disaster Risk Reduct.69, 102744. 10.1016/j.ijdrr.2021.102744

  • CrossRef
  • Google Scholar

• 84

  MorlighemM.RignotE.SeroussiH.LarourE.Ben DhiaH.AubryD. (2011). A mass conservation approach for mapping glacier ice thickness. Geophys. Res. Lett.38 (19). 10.1029/2011gl048659

  • CrossRef
  • Google Scholar

• 85

  MukherjeeK.BhattacharyaA.PieczonkaT.GhoshS.BolchT. (2018). Glacier mass budget and climate reanalysis data indicate a climatic shift around 2000 in Lahaul-Spiti, western Himalaya. Clim. Change148, 219–233. 10.1007/s10584-018-2185-3

  • CrossRef
  • Google Scholar

• 86

  NegiV. S.TiwariD. C.SinghL.ThakurS.BhattI. D. (2022). Review and synthesis of climate change studies in the Himalayan region. Environ. Dev. Sustain.24 (9), 10471–10502. 10.1007/s10668-021-01880-5

  • CrossRef
  • Google Scholar

• 87

  NieY.ShengY.LiuQ.LiuL.LiuS.ZhangY.et al (2017). A regional-scale assessment of Himalayan glacial lake changes using satellite observations from 1990 to 2015. Remote Sens. Environ.189, 1–13. 10.1016/j.rse.2016.11.008

  • CrossRef
  • Google Scholar

• 88

  NyeJ. F. (1952). The mechanics of glacier flow. J. Glaciol.2 (12), 82–93. 10.3189/s0022143000033967

  • CrossRef
  • Google Scholar

• 89

  PanditA.RamsankaranR. A. A. J. (2020). Modeling ice thickness distribution and storage volume of glaciers in Chandra Basin, western Himalayas. J. Mt. Sci.17 (8), 2011–2022. 10.1007/s11629-019-5718-y

  • CrossRef
  • Google Scholar

• 90

  PatelD. P.RamirezJ. A.SrivastavaP. K.BrayM.HanD. (2017). Assessment of flood inundation mapping of Surat city by coupled 1D/2D hydrodynamic modeling: a case application of the new HEC-RAS 5. Nat. Hazards89, 93–130. 10.1007/s11069-017-2956-6

  • CrossRef
  • Google Scholar

• 91

  PatersonW. S. B. (1970). The sliding velocity of Athabasca glacier, Canada. J. Glaciol.9 (55), 55–63. 10.3189/s0022143000026794

  • CrossRef
  • Google Scholar

• 92

  PatersonW. S. B. (1994). Physics of glaciers. Butterworth-Heinemann.

  • Google Scholar

• 93

  PaulF.AndreassenL. M. (2009). Creating a glacier inventory for the Svartisen region (Norway) from Landsat ETM+ satellite data: challenges and results. J. Glaciol.55, 607–618. 10.3189/002214309789471003

  • CrossRef
  • Google Scholar

• 94

  PaulF.LinsbauerA. (2012). Modeling of glacier bed topography from glacier outlines, central branch lines, and a DEM. Int. J. Geogr. Inf. Sci.26 (7), 1173–1190. 10.1080/13658816.2011.627859

  • CrossRef
  • Google Scholar

• 95

  PaulF.BarrandN. E.BaumannS.BerthierE.BolchT.CaseyK.et al (2013). On the accuracy of glacier outlines derived from remote-sensing data. Ann. Glaciol.54 (63), 171–182.

  • Google Scholar

• 96

  PaulF.WinsvoldS. H.KääbA.NaglerT.SchwaizerG. (2016). Glacier remote sensing using Sentinel-2. Part II: Mapping glacier extents and surface facies, and comparison to Landsat 8. Remote Sens.8 (7), 575.

  • Google Scholar

• 97

  PopovN. (1991). “Assessment of glacial debris flow hazard in the north Tien-Shan,” in Proceedings of the SovietChina-Japan Symposium and field workshop on natural disasters, September 2–7, 1991, 384–391.

  • Google Scholar

• 98

  PrakashC.NagarajanR. (2017). Outburst susceptibility assessment of moraine‐dammed lakes in Western Himalaya using an analytic hierarchy process. Earth Surface Processes Landforms42 (14), 2306–2321. 10.1002/esp.4185

  • CrossRef
  • Google Scholar

• 99

  QiW.FengL.KuangX.ZhengC.LiuJ.ChenD.et al (2022). Divergent and changing importance of glaciers and snow as natural water reservoirs in the eastern and southern Tibetan Plateau. J. Geophys. Res. Atmos.127 (7), e2021JD035888. 10.1029/2021jd035888

  • CrossRef
  • Google Scholar

• 100

  QuinceyD.LuckmanA.HesselR.DaviesR.SankhayanP.BallaM. (2007). Fine-resolution remote-sensing and modelling of Himalayan catchment sustainability. Remote Sens. Environ.107 (3), 430–439. 10.1016/j.rse.2006.09.021

  • CrossRef
  • Google Scholar

• 101

  RaiS. P.ThayyenR. J.PurushothamanP.KumarB. (2016). Isotopic characteristics of cryospheric waters in parts of Western Himalayas, India. Environ. Earth Sci.75, 1–9. 10.1007/s12665-016-5417-8

  • CrossRef
  • Google Scholar

• 102

  RamsankaranR. A. A. J.PanditA.AzamM. F. (2018). Spatially distributed ice-thickness modelling for Chhota Shigri Glacier in western Himalayas, India. Int. J. Remote Sens.39 (10), 3320–3343. 10.1080/01431161.2018.1441563

  • CrossRef
  • Google Scholar

• 103

  RatherA. F.AhmedR.AhmedP.BansalJ. K.BanerjeeP.WaniT. A.et al (2024a). Examining the glacier-glacial lake interactions of potentially dangerous glacial lakes (PDGLs) under changing climate in Shyok catchment of the Upper Indus Basin. Phys. Chem. Earth, Parts A/b/c136, 103686. 10.1016/j.pce.2024.103686

  • CrossRef
  • Google Scholar

• 104

  RatherA. F.AhmedR.JunC.BateniS. M.AhmedP.MirR. A. (2024b). Understanding glacial lake evolution and the associated GLOF hazard in the Shyok catchment of the Upper Indus Basin using geospatial techniques. Nat. Hazards Rev.25 (3), 04024014. 10.1061/nhrefo.nheng-1988

  • CrossRef
  • Google Scholar

• 105

  ReynoldsJ. M. (2000). On the Formation of supraglacial Lakes on Debris- covered glaciers, 153–161.

  • Google Scholar

• 106

  RichardsonS. D.ReynoldsJ. M. (2000). An overview of glacial hazards in the Himalayas. Quat. Int.65, 31–47. 10.1016/s1040-6182(99)00035-x

  • CrossRef
  • Google Scholar

• 107

  RounceD. R.HockR.MaussionF.HugonnetR.KochtitzkyW.HussM.et al (2023). Global glacier change in the 21st century: every increase in temperature matters. Science379 (6627), 78–83. 10.1126/science.abo1324

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 108

  SaatyT. L. (1980). The analytic hierarchy process. New York: McGraw Hill.

  • Google Scholar

• 109

  SattarA.GoswamiA.KulkarniA. V. (2019a). Application of 1D and 2D hydrodynamic modeling to study glacial lake outburst flood (GLOF) and its impact on a hydropower station in Central Himalaya. Nat. Hazards97 (2), 535–553. 10.1007/s11069-019-03657-6

  • CrossRef
  • Google Scholar

• 110

  SattarA.GoswamiA.KulkarniA. V.DasP. (2019b). Glacier-surface velocity derived ice volume and retreat assessment in the dhauliganga basin, central himalaya–A remote sensing and modeling-based approach. Front. Earth Sci.7, 105. 10.3389/feart.2019.00105

  • CrossRef
  • Google Scholar

• 111

  ScherlerD.BookhagenB.StreckerM. R. (2011). Spatially variable response of Himalayan glaciers to climate change affected by debris cover. Nat. Geoscience4 (3), 156–159. 10.1038/ngeo1068

  • CrossRef
  • Google Scholar

• 112

  SchmidtS.NüsserM. (2024). “Glaciers of Central Ladakh: distribution, changes and relevance in the Indian trans-Himalaya,” in Environmental change and development in Ladakh, Indian trans-himalaya (Cham: Springer Nature Switzerland), 11–30.

  • Google Scholar

• 113

  SchmidtS.NüsserM.BaghelR.DameJ. (2020). Cryosphere hazards in Ladakh: the 2014 Gya glacial lake outburst flood and its implications for risk assessment. Nat. Hazards104 (3), 2071–2095. 10.1007/s11069-020-04262-8

  • CrossRef
  • Google Scholar

• 114

  ShahabiH.HashimM.AhmadB. B. (2015). Remote sensing and GIS-based landslide susceptibility mapping using frequency ratio, logistic regression, and fuzzy logic methods at the central Zab basin, Iran. Environ. Earth Sci.73 (12), 8647–8668. 10.1007/s12665-015-4028-0

  • CrossRef
  • Google Scholar

• 115

  SharmaS.ShuklaA. D. (2018). Factors governing the pattern of glacier advances since the last Glacial Maxima in the transitional climate zone of the Southern Zanskar Ranges, NW Himalaya. Quat. Sci. Rev.201, 223–240. 10.1016/j.quascirev.2018.10.006

  • CrossRef
  • Google Scholar

• 116

  SheaJ. M.WagnonP.ImmerzeelW. W.BironR.BrunF.PellicciottiF. (2015). A comparative high-altitude meteorological analysis from three catchments in the Nepalese Himalaya. Int. Journal Water Resources Development31 (2), 174–200. 10.1080/07900627.2015.1020417

  • CrossRef
  • Google Scholar

• 117

  SheanD. E.BhushanS.MontesanoP.RounceD. R.ArendtA.OsmanogluB. (2020). A systematic, regional assessment of high mountain Asia glacier mass balance. Front. Earth Sci.7, 363. 10.3389/feart.2019.00363

  • CrossRef
  • Google Scholar

• 118

  ShugarD. H.BurrA.HaritashyaU. K.KargelJ. S.WatsonC. S.KennedyM. C.et al (2020). Rapid worldwide growth of glacial lakes since 1990. Nat. Clim. Chang.10 (10), 939–945. 10.1038/s41558-020-0855-4

  • CrossRef
  • Google Scholar

• 119

  SinghH.VaradeD.de VriesM. V. W.AdhikariK.RawatM.AwasthiS.et al (2023). Assessment of potential present and future glacial lake outburst flood hazard in the Hunza valley. A case study of Shisper and Mochowar glacier. Sci. Total Environment868, 161717. 10.1016/j.scitotenv.2023.161717

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 120

  SohebM.RamanathanA.BhardwajA.ColemanM.ReaB. R.SpagnoloM.et al (2022). Multitemporal glacier inventory revealing four decades of glacier changes in the Ladakh region. Earth Syst. Sci. Data14 (9), 4171–4185. 10.5194/essd-14-4171-2022

  • CrossRef
  • Google Scholar

• 121

  TakS.KeshariA. K. (2020). Investigating mass balance of Parvati glacier in Himalaya using satellite imagery-based model. Sci. Rep.10 (1), 12211. 10.1038/s41598-020-69203-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 122

  TaloorA. K.KothyariG. C.ManhasD. S.BishtH.MehtaP.SharmaM.et al (2021). Spatio-temporal changes in the Machoi glacier Zanskar Himalaya India using geospatial technology. Quat. Sci. Adv.4, 100031. 10.1016/j.qsa.2021.100031

  • CrossRef
  • Google Scholar

• 123

  TawdeS. A.KulkarniA. V.BalaG. (2019). An assessment of climate change impacts on glacier mass balance and geometry in the Chandra Basin, Western Himalaya for the 21st century. Environ. Res. Commun.1 (4), 041003. 10.1088/2515-7620/ab1d6d

  • CrossRef
  • Google Scholar

• 124

  TaylorC.RobinsonT. R.DunningS.Rachel CarrJ.WestobyM. (2023). Glacial lake outburst floods threaten millions globally. Nat. Communications14 (1), 487. 10.1038/s41467-023-36033-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 125

  ThayyenR. J.MishraP. K.JainS. K.WaniJ. M.SinghH.SinghM. K.et al (2022). Hanging glacier avalanche (Raunthigad–Rishiganga) and debris flow disaster on 7 February 2021, Uttarakhand, India. A preliminary assessment. Nat. Hazards114 (2), 1939–1966. 10.1007/s11069-022-05454-0

  • CrossRef
  • Google Scholar

• 126

  Van TrichtL.PaiceC. M.RybakO.HuybrechtsP. (2023). Brief communication: measuring and modelling the ice thickness of the Grigoriev ice cap (Kyrgyzstan) and comparison with global datasets. Cryosphere17 (10), 4315–4323. 10.5194/tc-17-4315-2023

  • CrossRef
  • Google Scholar

• 127

  Van Wyk de VriesM.Carchipulla-MoralesD.WickertA. D.MinayaV. G. (2022). Glacier thickness and ice volume of the Northern Andes. Sci. Data9 (1), 342. 10.1038/s41597-022-01446-8

  • CrossRef
  • Google Scholar

• 128

  VehG.KorupO.RoessnerS.WalzA. (2018). Detecting Himalayan glacial lake outburst floods from Landsat time series. Remote Sens. Environ.207, 84–97. 10.1016/j.rse.2017.12.025

  • CrossRef
  • Google Scholar

• 129

  VehG.KorupO.SpechtS.RoessnerS.WalzA. (2019). Unchanged frequency of moraine-dammed glacial lake outburst floods in the Himalaya. Nat. Clim. Chang.9 (5), 379–383. 10.1038/s41558-019-0437-5

  • CrossRef
  • Google Scholar

• 130

  VehG.LützowN.KharlamovaV.PetrakovD.HugonnetR.KorupO. (2022). Trends, breaks, and biases in the frequency of reported Glacier Lake outburst floods. Earth's Future10 (3), e2021EF002426. 10.1029/2021EF002426

  • CrossRef
  • Google Scholar

• 131

  WangW.XiangY.GaoY.LuA.YaoT. (2015). Rapid expansion of glacial lakes caused by climate and glacier retreat in the Central Himalayas. Hydrol. Process.29 (6), 859–874. 10.1002/hyp.10199

  • CrossRef
  • Google Scholar

• 132

  WashakhR. M. A.ChenN.WangT.AlmasS.AhmadS. R.RahmanM. (2019). GLOF risk assessment model in the himalayas: a case study of a hydropower project in the upper arun river. Water11 (9), 1839. 10.3390/w11091839

  • CrossRef
  • Google Scholar

• 133

  WestobyM. J.GlasserN. F.HambreyM. J.BrasingtonJ.ReynoldsJ. M.HassanM. A. A. M. (2014a). Reconstructing historic glacial Lake outburst floods through numerical modelling and geomorphological assessment. Extreme events in the Himalaya. Earth Surf. Process. Landf.39 (12), 1675–1692. 10.1002/esp.3617

  • CrossRef
  • Google Scholar

• 134

  WestobyM. J.GlasserN. F.BrasingtonJ.HambreyM. J.QuinceyD. J.ReynoldsJ. M. (2014b). Modelling outburst floods from moraine-dammed glacial lakes. Earth-Science Rev.134, 137–159. 10.1016/j.earscirev.2014.03.009

  • CrossRef
  • Google Scholar

• 135

  WorniR.HuggelC.ClagueJ. J.SchaubY.StoffelM. (2014). Coupling glacial lake impact, dam breach, and flood processes: a modeling perspective. Geomorphology224, 161–176. 10.1016/j.geomorph.2014.06.031

  • CrossRef
  • Google Scholar

• 136

  YaoX.LiuS.HanL.SunM.ZhaoL. (2018). Definition and classification system of glacial lake for inventory and hazards study. J. Geogr. Sci.28, 193–205. 10.1007/s11442-018-1467-z

  • CrossRef
  • Google Scholar

• 137

  ZempM.ThibertE.HussM.StummD.Rolstad DenbyC.NuthC.et al (2013). Reanalysing glacier mass balance measurement series. Cryosphere7 (4), 1227–1245. 10.5194/tc-7-1227-2013

  • CrossRef
  • Google Scholar

• 138

  ZhangG.YaoT.XieH.WangW.YangW. (2015). An inventory of glacial lakes in the third pole region and their changes in response to global warming. Glob. Planet. Change131, 148–157. 10.1016/j.gloplacha.2015.05.013

  • CrossRef
  • Google Scholar

• 139

  ZhangM.ChenF.ZhaoH.WangJ.WangN. (2021). Recent changes of glacial Lakes in the high Mountain Asia and its potential controlling factors analysis. Remote Sens.13 (18), 3757. 10.3390/rs13183757

  • CrossRef
  • Google Scholar

• 140

  ZhangT.WangW.GaoT.AnB.YaoT. (2022). An integrative method for identifying potentially dangerous glacial lakes in the Himalayas. Sci. Total Environ.806, 150442. 10.1016/j.scitotenv.2021.150442

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 141

  ZhangG.BolchT.YaoT.RounceD. R.ChenW.VehG.et al (2023). Underestimated mass loss from lake-terminating glaciers in the greater Himalaya. Nat. Geosci.16 (4), 333–338. 10.1038/s41561-023-01150-1

  • CrossRef
  • Google Scholar

• 142

  ZhouY.LiZ.LiJ.ZhaoR.DingX. (2018). Glacier mass balance in the Qinghai–Tibet Plateau and its surroundings from the mid-1970s to 2000 based on Hexagon KH-9 and SRTM DEMs. Remote Sens. Environ.210 (1), 96–112. 10.1016/j.rse.2018.03.020

  • CrossRef
  • Google Scholar