High-resolution Snow Depth Research Articles

A combined data assimilation and deep learning approach for continuous spatio-temporal SWE reconstruction from sparse ground tracks

Our understanding of the impact of climate change on water availability and natural hazards in high-mountain regions is limited due to the spatial and temporal scarcity of ground observations of precipitation and snow. Freely available, satellite-based information about the snowpack is currently mainly limited to indirect measurements of snow-covered area or very coarse-scale snow water equivalent (SWE), but only for flat areas in lowlands without vegetation cover. Novel space-based laser altimeters, such as ICESat-2, have the potential to provide high-resolution snow depth data in worldwide mountain regions where no ground observations exist. However, these space-based laser altimeters come with spatial gaps between ground tracks, obtained without repetition at a give location. To overcome these drawbacks, here, we present a combined probabilistic data assimilation and deep learning approach to reconstruct spatio-temporal SWE from observations of snow depth along ground tracks, imitating ICESat-2 tracks in view of a potential future global application.Our approach is based on assimilating SWE and snow cover information in a degree-day model with an iterative ensemble smoother (IES) which allows temporally reconstructing SWE along hypothetical ground tracks separated by 3 km. As input, the degree-day model uses daily precipitation and downscaled air temperature from the ERA5 reanalysis. A feedforward neural network (FNN) is then used for spatial propagation of the daily mean and standard deviation of the updated SWE ensemble members obtained from the IES. The combined IES-FNN approach provides uncertainty-aware spatio-temporally continuous estimates of SWE.We tested our approach in the alpine Dischma valley (Switzerland) using high-resolution snow depth maps obtained from photogrammetric techniques mounted on airplanes and unmanned aerial system observations. Our results show that the IES-FNN model provides reliable estimates at a resolution of approximately 100 m. Even assimilating only one SWE observation during the year (combined with satellite-based melt-out date estimates) produces satisfying results when evaluating the IES-FNN SWE reconstructions on independent dates and smaller (<4 km2) areas: mean absolute error of 86 mm (78 mm) at Schürlialp (Latschüelfurgga) for average SWE of 180 mm (254 mm), and average spatial linear correlation with the reference SWE of 0.51 (0.48). However, the assimilated SWE observation must not be too early in the accumulation season or too late in the melt season when the snowpack is starting or ending to accumulate or melt, respectively. Smaller distances between ground tracks (1500 m and 500 m) show improved performance of the IES-FNN approach in space, with no significant improvement in terms of temporal reconstruction.Applying the IES-FNN approach to e.g., real ICESat-2 data, remains challenging due to the higher uncertainties associated with these data. However, the approach we propose remains potentially very helpful in addressing the problem of scarcity of ground observations of precipitation and snow in high-mountain regions.

Estimating high-resolution snow depth over the North Hemisphere mountains utilizing active microwave backscatter and machine learning

While ground meteorological stations provide accurate snow depth data, their limited spatial coverage results in observational gaps. Satellites offer long-term, large-scale observations, addressing these gaps. Existing snow depth retrieval algorithms mainly use passive microwave remote sensing data with a 25 km resolution, insufficient for capturing snow depth variability in mountainous areas. This paper introduces active microwave backscatter data and machine learning techniques for high-resolution snow depth estimation. We conducted a preliminary exploration of the relationship between Sentinel-1 backscatter coefficient σ0 and snow depth. Due to factors such as vegetation coverage and underlying soil properties, the relationship between σ0 and snow depth is complex and nonlinear. Consequently, six machine learning models were trained to learn this relationship using σ0 and auxiliary data as input features, with in-situ snow depth serving as the target variable. After extensive validation, the Extreme Random Trees (ERT) model was selected for its high accuracy and stability. Using the ERT model, we generated 500 m-resolution snow depth data for northern hemisphere mountains, then analyzed temporal snow depth variations and altitudinal stratification.

Passive microwave remote-sensing-based high-resolution snow depth mapping for Western Himalayan zones using multifactor modeling approach

Abstract. Spatiotemporal snow depth (SD) mapping in the Indian Western Himalayan (WH) region is essential in many applications pertaining to hydrology, natural disaster management, climate, etc. In situ techniques for SD measurement are not sufficient to represent the high spatiotemporal variability in SD in the WH region. Currently, low-frequency passive microwave (PMW) remote-sensing-based algorithms are extensively used to monitor SD at regional and global scales. However, fewer PMW SD estimation studies have been carried out for the WH region to date, which are mainly confined to small subregions of the WH region. In addition, the majority of the available PMW SD models for WH locations are developed using limited data and fewer parameters and therefore cannot be implemented for the entire region. Further, these models have not taken the auxiliary parameters such as location, topography, and snow cover duration (SCD) into consideration and have poor accuracy (particularly in deep snow) and coarse spatial resolution. Considering the high spatiotemporal variability in snow depth characteristics across the WH region, region-wise multifactor models are developed for the first time to estimate SD at a high spatial resolution of 500 m × 500 m for three different WH zones, i.e., Lower Himalayan Zone (LHZ), Middle Himalayan Zone (MHZ), and Upper Himalayan Zone (UHZ). Multifrequency brightness temperature (TB) observations from Advanced Microwave Scanning Radiometer 2 (AMSR2), SCD data, terrain parameters (i.e., elevation, slope, and ruggedness), and geolocation for the winter period (October to March) during 2012–2013 to 2016–2017 are used for developing the SD models for dry snow conditions. Different regression approaches (i.e., linear, logarithmic, reciprocal, and power) are used to develop snow depth models, which are evaluated further to find if any of these models can address the heterogeneous association between SD observations and PMW TB. From the results, it is observed from the analysis that the power regression SD model has improved accuracy in all WH zones with the low root mean square error (RMSE) in the MHZ (i.e., 27.21 cm) compared to the LHZ (32.87 cm) and the UHZ (42.81 cm). The spatial distribution of model-derived SD is highly affected by SCD, terrain parameters, and geolocation parameters and has better SD estimates compared to regional and global products in all zones. Overall results indicate that the proposed multifactor SD models have achieved higher accuracy in deep snowpack (i.e., SD &gt;25 cm) of the WH region compared to previously developed SD models.

Snow accumulation, albedo and melt patterns following road construction on permafrost, Inuvik–Tuktoyaktuk Highway, Canada

Abstract. Roads constructed on permafrost can have a significant impact on the surrounding environment, potentially inducing permafrost degradation. These impacts arise from factors such as snow accumulation near the road, which affects the soil's thermal and hydrological regime, and road dust that decreases the snow's albedo, altering the timing of snowmelt. However, our current understanding of the magnitude and the spatial extent of these effects is limited. In this study we addressed this gap by using remote sensing techniques to assess the spatial effect of the Inuvik to Tuktoyaktuk Highway (ITH) in Northwest Territories, Canada, on snow accumulation, snow albedo and snowmelt patterns. With a new, high resolution snow depth raster from airborne laser scanning, we quantified the snow accumulation at road segments in the Trail Valley Creek area using digital elevation model differencing. We found increased snow accumulation up to 36 m from the road center. The magnitude of this snow accumulation was influenced by the prevailing wind direction and the embankment height. Furthermore, by analyzing 43 Sentinel-2 satellite images between February and May 2020, we observed reduced snow albedo values within 500 m of the road, resulting in a 12-days-earlier onset of snowmelt within 100 m from the road. We examined snowmelt patterns before, during and after the road construction using the normalized difference snow index from Landsat-7 and Landsat-8 imagery. Our analysis revealed that the road affected the snowmelt pattern up to 600 m from the road, even in areas which appeared undisturbed. In summary, our study improves our understanding of the spatial impact of gravel roads on permafrost due to enhanced snow accumulation, reduced snow albedo and earlier snowmelt. Our study underscores the important contribution that remote sensing can provide to improve our understanding of the effects of infrastructure development on permafrost environments.

Measuring the spatiotemporal variability in snow depth in subarctic environments using UASs – Part 2: Snow processes and snow–canopy interactions

Abstract. Detailed information on seasonal snow cover and depth is essential to the understanding of snow processes, to operational forecasting, and as input for hydrological models. Recent advances in uncrewed or unmanned aircraft systems (UASs) and structure from motion (SfM) techniques have enabled low-cost monitoring of spatial snow depth distribution in resolutions of up to a few centimeters. Here, we study the spatiotemporal variability in snow depth and interactions between snow and vegetation in different subarctic landscapes consisting of a mosaic of conifer forest, mixed forest, transitional woodland/shrub, and peatland areas. To determine the spatiotemporal variability in snow depth, we used high-resolution (50 cm) snow depth maps generated from repeated UAS–SfM surveys in the winter of 2018/2019 and a snow-free bare-ground survey after snowmelt. Due to poor subcanopy penetration with the UAS–SfM method, tree masks were utilized to remove canopy areas and the area (36 cm) immediately next to the canopy before analysis. Snow depth maps were compared to the in situ snow course and a single-point continuous ultrasonic snow depth measurement. Based on the results, the difference between the UAS–SfM survey median snow depth and single-point measurement increased for all land cover types during the snow season, from +5 cm at the beginning of the accumulation to −16 cm in coniferous forests and −32 cm in peatland during the melt period. This highlights the poor representation of point measurements in selected locations even on the subcatchment scale. The high-resolution snow depth maps agreed well with the snow course measurement, but the spatial extent and resolution of maps were substantially higher. The snow depth range (5th–95th percentiles) within different land cover types increased from 17 to 42 cm in peatlands and from 33 to 49 cm in the coniferous forest from the beginning of the snow accumulation to the melt period. Both the median snow depth and its range were found to increase with canopy density; this increase was greatest in the conifer forest area, followed by mixed forest, transitional woodland/shrub, and open peatlands. Using the high-spatial-resolution data, we found a systematic increase (2–20 cm) and then a decline in snow depth near the canopy with increasing distance (from 1 to 2.5 m) of the peak value through the snow season. This study highlights the applicability of the UAS–SfM in high-resolution monitoring of snow depth in multiple land cover types and snow–vegetation interactions in subarctic and remote areas where field data are not available or where the available data are collected using classic point measurements or snow courses.

Snow depth estimation at country-scale with high spatial and temporal resolution

Monitoring snow depth is important for applications such as hydrology, energy planning, ecology, and safety evaluation for outdoor winter activities. Most methods able to estimate snow depth for large regions can only do so in a spatial resolution of up to 1 km ground sampling distance (GSD). This limits their usage in high alpine areas, where this resolution fails to capture local snow distribution patterns caused by the pronounced topographical features. In this work we use a recurrent convolutional neural network to estimate snow depth at high spatial resolution (10 m GSD), weekly, and at large scale based on satellite data sources and elevation maps, without the need for measurement stations on the ground. The proposed method achieves unprecedented results for large-scale, high-resolution snow depth mapping. The resulting maps are evaluated over a period of three years against high-fidelity snow depth maps obtained with airborne photogrammetry. Finally, we also produce well-calibrated uncertainty estimates for every individual snow depth estimate via a probabilistic regression framework.

Machine Learning-Based Estimation of High-Resolution Snow Depth in Alaska Using Passive Microwave Remote Sensing Data

Machine Learning-Based Estimation of High-Resolution Snow Depth in Alaska Using Passive Microwave Remote Sensing Data

Applicability of alpine snow depth estimation based on multitemporal UAV-LiDAR data: A case study in the Maxian Mountains, Northwest China

Unmanned aerial vehicle (UAV) LiDAR has emerged as a viable approach to generate high-resolution snow depth (SD) maps. However, implementing this technology in high-elevation mountains is difficult, and few studies have examined the SD on mountains in Asia until now. Here, we used UAV-LiDAR to estimate the SD on the Maxian Mountains along the eastern edge of High Mountain Asia to assess the reliability of this technology for obtaining SD distributions on a typical alpine periglacial geomorphology. Using LiDAR data from different periods (i.e., 2020/10/19, 2020/11/25 and 2021/04/08) and evenly-distributed in situ measurements, we obtained and validated SDs for both shallow (mean SD less than 15 cm) and deep (mean SD ≥ 15 cm) snow conditions. Our results indicated that the quality of the LiDAR point cloud data met the requirements of the relevant specifications, and the SD distribution map captured the spatial heterogeneity caused by the hummocks. The multitemporal SDs derived from LiDAR were all overestimated, but the accuracy for shallow snow was better than that for deep snow. The root mean square error (RMSE) and mean absolute error (MAE) are 2.23 cm and 1.57 cm, respectively, for deep snow, which were almost three times those of shallow snow. The seasonal deformation of the hummock and manual sampling errors caused by the ice layer over the soil were the main factors affecting accuracy. In contrast, snow surface conditions played a minor role in the error estimators. This study helps refine the use of UAV-LiDAR for SD estimation in similar environments.

Mapping snow depth and spatial variability using SFM photogrammetry of UAV images over rugged mountainous regions of the Western Himalaya

Accurate snow depth observations are one of the most important inputs in the process of avalanche hazard estimation, forecasting, and mitigation. This study presents the high-resolution snow depth mapping of avalanche-prone regions of the Manali–Dhundi area, Himachal Pradesh (HP), India, using, repeated UAV Photogrammetric surveys. The images captured by an RGB camera mounted on a fixed-wing UAV (e-Bee X) were used to generate the high-resolution Digital Surface Model (DSM) and Orthophotos of the study area. Mapping of snow depth of the study area is crucial for avalanche hazard assessment and safe movement of the personnel and vehicles passing through the highways in this area during winters. A total of 12 km2 area covering six major avalanche sites was surveyed. A snow depth map was generated by differencing co-registered snow-covered DSM and bare surface DSM. UAV-derived snow depth values have been validated with the point snow depth measurements of Wireless Sensor Networks (WSN), manual observation station, and snow depth around a structure of known height in the survey area. The UAV-retrieved snow depth values were found in good correlation with field-measured snow depth (R = 0.95). The maximum and minimum observed snow depth values at different locations on open slopes and flat areas were 1.62 m and 0.79 m. The estimated Mean Error (ME) and RMSE based on these measured values were found 0.15 and 0.21 m, respectively. The study reveals that the snow depth map generated from the UAV survey provides reliable estimates of snow depth values for flat areas and open slopes but the method is not suitable for forested areas. The study also infers that the statistical measures (i.e., µ, σ, RMSE, etc.) derived from the complete survey dataset without considering topographic factors and features of the region will not provide realistic estimates of snow depth and its accuracy. The study suggests that regular monitoring (before and after every snow storm) of the snow depth using a UAV-based method is quick, cost-effective, and provides accurate snow conditions for operational use and planning in snow-bound regions.

High-Resolution Snow Depth on Arctic Sea Ice From Low-Altitude Airborne Microwave Radar Data

We present new high-resolution snow depth data on Arctic sea ice derived from airborne microwave radar measurements from the IceBird campaigns of the Alfred Wegener Institute (AWI) together with a new retrieval method using signal peakiness based on an intercomparison exercise of colocated data at different altitudes. We aim to demonstrate the capabilities and potential improvements of radar data, which were acquired at a lower altitude (200 ft) and slower speed (110 kn) and had a smaller radar footprint size (2-m diameter) than previous airborne snow radar data. So far, AWI Snow Radar data have been derived using a 2–18-GHz ultrawideband frequency-modulated continuous-wave (FMCW) radar in 2017–2019. Our results show that our method in combination with thorough calibration through coherent noise removal and system response deconvolution significantly improves the quality of the radar-derived snow depth data. The validation against a 2-D grid of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in situ</i> snow depth measurements on level landfast first-year ice indicates a mean bias of only 0.86 cm between radar and ground truth. Comparison between the radar-derived snow depth estimates from different altitudes shows good consistency. We conclude that the AWI Snow Radar aboard the IceBird campaigns is able to measure the snow depth on Arctic sea ice accurately at higher spatial resolution than but consistent with the existing airborne snow radar data of NASA Operation IceBridge. Together with the simultaneous measurements of the total ice thickness and surface freeboard, the IceBird campaign data will be able to describe the whole sea-ice column on regional scales.

Unpiloted Aerial Vehicle Retrieval of Snow Depth Over Freshwater Lake Ice Using Structure From Motion

The presence and thickness of snow overlying lake ice affects both the timing of melt and ice-free conditions, can contribute to overall ice thickness through its insulative capacity, and fosters the development of variable ice types. The use of UAVs to retrieve snow depths with high spatial resolution is necessary for the next generation of ultra-fine hydrological models, as the direct contribution of water from snow on lake ice is unknown. Such information is critical to the understanding of the physical processes of snow redistribution and capture in catchments on small lakes in the Arctic, which has been historically estimated from its relationship to terrestrial snowpack properties. In this study, we use a quad-copter UAV and SfM principles to retrieve and map snow depth at the winter maximum at high resolution over a the freshwater West Twin Lake on the Arctic Coastal Plain of northern Alaska. The accuracy of the snow depth retrievals is assessed using in-situ observations (n = 1,044), applying corrections to account for the freeboard of floating ice. The average snow depth from in-situ observations was used calculate a correction factor based on the freeboard of the ice to retrieve snow depth from UAV acquisitions (RMSE = 0.06 and 0.07 m for two transects on the lake. The retrieved snow depth map exhibits drift structures that have height deviations with a root mean square (RMS) of 0.08 m (correlation length = 13.8 m) for a transect on the west side of the lake, and an RMS of 0.07 m (correlation length = 18.7 m) on the east. Snow drifts present on the lake also correspond to previous investigations regarding the variability of snow on lakes, with a periodicity (separation) of 20 and 16 m for the west and east side of the lake, respectively. This study represents the first retrieval of snow depth on a frozen lake surface from a UAV using photogrammetry, and promotes the potential for high-resolution snow depth retrieval on small ponds and lakes that comprise a significant portion of landcover in Arctic environments.

Accounting for permafrost creep in high-resolution snow depth mapping by modelling sub-snow ground deformation

Snow depth estimation derived from high-resolution digital elevation models (DEMs) can lead to improved understanding of the spatially highly heterogeneous nature of snow distribution, as well as help us improve our knowledge of how snow patterns influence local geomorphic processes. Slope deformation processes such as permafrost creep can make it challenging to acquire a snow-free DEM that matches the sub-snow topography at the time of the associated snow-covered DEM, which can cause errors in the computed snow depths. In this study, we illustrate how modelling changes in the sub-snow topography can reduce errors in snow depths derived from DEM differencing in an area of permafrost creep. To model the sub-snow topography, a surface deformation model was constructed by performing non-rigid registration based on B-splines of two snow-free DEMs. Seasonal variations in creep were accounted for by using an optimization approach to find a suitable value to scale the deformation model based on in-situ snow depth measurements or the presence of snow-free areas corresponding to the date of the snow-covered DEM. This scaled deformation model was used to transform one of the snow-free DEMs to estimate the sub-snow topography corresponding to the date of the snow-covered DEM. The performance of this method was tested on an active rock glacier in the southern French Alps for two surveys dates, which were conducted in the winter and spring of 2017.By accounting for surface displacements caused by permafrost creep, we found that our method was able to reduce the errors in the estimated snow depths by up to 33% (an interquartile range reduction of 11 cm) compared to using the untransformed snow-free DEM. The accuracy of the snow depths only slightly improved (root-mean-square error decrease of up to 3 cm). Greater reductions in error were observed for the snow depths calculated for the date that was furthest (i.e., the winter survey) in time from the snow-free DEM. Additionally, we found that our approach to scaling the deformation model has promising potential to be adapted for monitoring seasonal variations in permafrost creep by combining in-situ snow depth measurements with high-resolution surface deformation models.

Application of Low-Cost UASs and Digital Photogrammetry for High-Resolution Snow Depth Mapping in the Arctic

The repeat acquisition of high-resolution snow depth measurements has important research and civil applications in the Arctic. Currently the surveying methods for capturing the high spatial and temporal variability of the snowpack are expensive, in particular for small areal extents. An alternative methodology based on Unmanned Aerial Systems (UASs) and digital photogrammetry was tested over varying surveying conditions in the Arctic employing two diverse and low-cost UAS-camera combinations (500 and 1700 USD, respectively). Six areas, two in Svalbard and four in Greenland, were mapped covering from 1386 to 38,410 m2. The sites presented diverse snow surface types, underlying topography and light conditions in order to test the method under potentially limiting conditions. The resulting snow depth maps achieved spatial resolutions between 0.06 and 0.09 m. The average difference between UAS-estimated and measured snow depth, checked with conventional snow probing, ranged from 0.015 to 0.16 m. The impact of image pre-processing was explored, improving point cloud density and accuracy for different image qualities and snow/light conditions. Our UAS photogrammetry results are expected to be scalable to larger areal extents. While further validation is needed, with the inclusion of extra validation points, the study showcases the potential of this cost-effective methodology for high-resolution monitoring of snow dynamics in the Arctic and beyond.

Distributed snow and rock temperature modelling in steep rock walls using Alpine3D

Abstract. In this study we modelled the influence of the spatially and temporally heterogeneous snow cover on the surface energy balance and thus on rock temperatures in two rugged, steep rock walls on the Gemsstock ridge in the central Swiss Alps. The heterogeneous snow depth distribution in the rock walls was introduced to the distributed, process-based energy balance model Alpine3D with a precipitation scaling method based on snow depth data measured by terrestrial laser scanning. The influence of the snow cover on rock temperatures was investigated by comparing a snow-covered model scenario (precipitation input provided by precipitation scaling) with a snow-free (zero precipitation input) one. Model uncertainties are discussed and evaluated at both the point and spatial scales against 22 near-surface rock temperature measurements and high-resolution snow depth data from winter terrestrial laser scans.In the rough rock walls, the heterogeneously distributed snow cover was moderately well reproduced by Alpine3D with mean absolute errors ranging between 0.31 and 0.81 m. However, snow cover duration was reproduced well and, consequently, near-surface rock temperatures were modelled convincingly. Uncertainties in rock temperature modelling were found to be around 1.6 °C. Errors in snow cover modelling and hence in rock temperature simulations are explained by inadequate snow settlement due to linear precipitation scaling, missing lateral heat fluxes in the rock, and by errors caused by interpolation of shortwave radiation, wind and air temperature into the rock walls.Mean annual near-surface rock temperature increases were both measured and modelled in the steep rock walls as a consequence of a thick, long-lasting snow cover. Rock temperatures were 1.3–2.5 °C higher in the shaded and sunny rock walls, while comparing snow-covered to snow-free simulations. This helps to assess the potential error made in ground temperature modelling when neglecting snow in steep bedrock.

Snow as a driving factor of rock surface temperatures in steep rough rock walls

Observations show that considerable amounts of snow can accumulate in steep, rough rock walls. The heterogeneously distributed snow cover significantly affects the surface energy balance and hence the thermal regime of the rock walls.To assess the small-scale variability of snow depth and rock temperatures in steep north and south facing rock walls, a spatially distributed multi-method approach is applied at Gemsstock, Switzerland, combining 35 continuous near-surface rock temperature measurements, high resolution snow depth observations using terrestrial laser scanning, as well as in-situ snow pit investigations.The thermal regime of the rock surface is highly dependent on short- and longwave radiation, albedo, surface roughness, snow depth and on snow distribution in time and space. Around 2m of snow can accumulate on slopes with angles up to 75°, due to micro-topographic structures like ledges. Hence, contrasts in mean annual rock surface temperature between the north and the south facing slopes are less than 4°C. However, significant small-scale variability of up to 10°C in mean daily rock surface temperature occurs within a few metres over the rock walls due to the variable snow distribution, revealing the heterogeneity and complexity of the thermal regime at a very local scale. In addition, multiple linear regression could explain up to 77% of near‐surface rock temperature variability, which underlines the importance of radiation and snow depth and thus also of the topography.In the rock faces the thermal insulation of the ground starts with snow depths exceeding 0.2m. This is due to the high thermal resistance of a less densely packed snow cover, especially in the north facing slope. Additionally, aspect induced differences of snow cover characteristics and consequently thermal conductivities are observed in the rock walls.

Methodological approaches to infer end-of-winter snow distribution on alpine glaciers

AbstractSnow accumulation is an important component of the mass balance of alpine glaciers. To improve our understanding of the processes related to accumulation and their representation in state-of-the-art mass-balance models, extensive field measurements are required. We present measurements of snow accumulation distribution on Findelengletscher, Switzerland, for April 2010 using (1) in situ snow probings, (2) airborne ground-penetrating radar (GPR) and (3) differencing of two airborne light detection and ranging (lidar) digital elevation models (DEMs). Calculating high-resolution snow depth from DEM-differencing requires careful correction for vertical ice-flow velocity and densification in the accumulation area. All three methods reveal a general increase in snow depth with elevation, but also a significant small-scale spatial variability. Lidar-differencing and in situ snow probings show good agreement for the mean specific winter balance (0.72 and 0.78 m w.e., respectively). The lidar-derived distributed snow depth reveals significant zonal correlations with elevation, slope and curvature in a multiple linear regression model. Unlike lidar-differencing, GPR-derived snow depth is not affected by glacier dynamics or firn compaction, but to a smaller degree by snow density and liquid water content. It is thus a valuable independent data source for validation. The simultaneous availability of the three datasets facilitates the comparison of the methods and contributes to a better understanding of processes that govern winter accumulation distribution on alpine glaciers.