Snow Depth Measurements Research Articles

Spatially Extensive Ground‐Penetrating Radar Snow Depth Observations During NASA's 2017 SnowEx Campaign: Comparison With In Situ, Airborne, and Satellite Observations

AbstractSeasonal snow is an important component of Earth's hydrologic cycle and climate system, yet it remains challenging to consistently and accurately measure snow depth and snow water equivalent (SWE) across the range of diverse snowpack conditions that exist on Earth. The NASA SnowEx campaign is focused on addressing the primary gaps in snow remote sensing in order to gain an improved spatiotemporal understanding of this important resource and to further efforts toward a future satellite‐based snow remote sensing mission. Ground‐penetrating radar (GPR) is an efficient and mature approach for measuring snow depth and SWE. We collected ~1.3 million GPR snow depth observations during the NASA SnowEx 2017 campaign, yielding a spatially extensive (~133‐km total length) and high‐resolution (~10‐cm lateral spacing) validation data set to assess various remote sensing and modeling approaches. We found high correlation between the GPR and manual snow probe derived snow depths (r = 0.89, p < 0.0001, root‐mean‐square error (RMSE) = 18 cm), but a median difference of −10 cm, which we attribute, in part, to probe penetration into the unfrozen subsurface. We also compared GPR‐derived snow depths to two other independent estimates of snow depth, as an example of how this data set can be used for validation of remote sensing techniques: Airborne Snow Observatory lidar‐derived snow depths (r = 0.90, p < 0.0001, median difference = −1 cm, RMSE = 14 cm) and preliminary DigitalGlobe WorldView‐3 satellite‐derived snow depths (r = 0.70, p < 0.0001, median difference = −3 cm, RMSE = 24 cm).

Estimating snow depth on Arctic sea ice using satellite microwave radiometry and a neural network

Abstract. Snow lying on top of sea ice plays an important role in the radiation budget because of its high albedo and the Arctic freshwater budget, and it influences the Arctic climate: it is a fundamental climate variable. Importantly, accurate snow depth products are required to convert satellite altimeter measurements of ice freeboard to sea ice thickness (SIT). Due to the harsh environment and challenging accessibility, in situ measurements of snow depth are sparse. The quasi-synoptic frequent repeat coverage provided by satellite measurements offers the best approach to regularly monitor snow depth on sea ice. A number of algorithms are based on satellite microwave radiometry measurements and simple empirical relationships. Reducing their uncertainty remains a major challenge. A High Priority Candidate Mission called the Copernicus Imaging Microwave Radiometer (CIMR) is now being studied at the European Space Agency. CIMR proposes a conically scanning radiometer having a swath >1900 km and including channels at 1.4, 6.9, 10.65, 18.7 and 36.5 GHz on the same platform. It will fly in a high-inclination dawn–dusk orbit coordinated with the MetOp-SG(B). As part of the preparation for the CIMR mission, we explore a new approach to retrieve snow depth on sea ice from multi-frequency satellite microwave radiometer measurements using a neural network approach. Neural networks have proven to reach high accuracies in other domains and excel in handling complex, non-linear relationships. We propose one neural network that only relies on AMSR2 channel brightness temperature data input and another one using both AMSR2 and SMOS data as input. We evaluate our results from the neural network approach using airborne snow depth measurements from Operation IceBridge (OIB) campaigns and compare them to products from three other established snow depth algorithms. We show that both our neural networks outperform the other algorithms in terms of accuracy, when compared to the OIB data and we demonstrate that plausible results are obtained even outside the algorithm training period and area. We then convert CryoSat freeboard measurements to SIT using different snow products including the snow depth from our networks. We confirm that a more accurate snow depth product derived using our neural networks leads to more accurate estimates of SIT, when compared to the SIT measured by a laser altimeter at the OIB campaign. Our network with additional SMOS input yields even higher accuracies, but has the disadvantage of a larger “hole at the pole”. Our neural network approaches are applicable over the whole Arctic, capturing first-year ice and multi-year ice conditions throughout winter. Once the networks are designed and trained, they are fast and easy to use. The combined AMSR2 + SMOS neural network is particularly important as a precursor demonstration for the Copernicus CIMR candidate mission highlighting the benefit of CIMR.

Fully automated snow depth measurements from time-lapse images applying a convolutional neural network

Time-lapse cameras in combination with simple measuring rods can form a highly reliable low-cost sensor network monitoring snow depth in a high spatial and temporal resolution. Depending on the number of cameras and the temporal recording resolution, such a network produces large sets of image time series. In order to extract the snow depth time series from these collections of images in acceptable time, automated processing methods have to be applied. Besides classic image processing based on edge detection methods, there are nowadays ready-to-use convolutional neural network frameworks like Mask R-CNN that facilitate instance segmentation and thus allow for fully automated snow depth measurements from images using a detectable measuring rod. This study investigates the applicability of Mask R-CNN embedded in a newly developed work flow for snow depth measurements. The new method is compared to an automated image processing method carried out utilizing functionalities provided by the OpenCV library. The quality of both methods was assessed with the inclusion of manual evaluations of the image series. As a result, the newly introduced work flow outperforms the present classic image processing method in regard to stability, accuracy and portability. By applying the Mask R-CNN framework, the overall RMSE of two considered time series is reduced to approximately 20% of the value produced by means of the classic image processing approach. Moreover, the ratio of values within five centimeter deviation from the reference value was increased from 75% to 88% on average. Since no parameters have to be adjusted, the Mask R-CNN framework is able to detect known shapes reliably in almost any environment, making the presented method highly flexible.

Estimating the snow water equivalent from snow depth measurements in the Italian Alps

The Snow Water Equivalent (SWE), combining the information of snow depth (Hs) and snow bulk density (ρb) is a necessary variable for snow-hydrological studies and applications, as well as, for ecological function or avalanche forecasting. The SWE direct measurement is challenging, and estimating the SWE from the single Hs measurements presents many advantages compared to the direct SWE measurement or the implementation of complex model needing to be fed by local meteorological data. In this study we propose a spatial and temporal variability description of the SWE, Hs and ρb and compare existing approaches over the Italian Alps. Finally, we propose a simple parametrization, introducing non-linearity in the snow bulk density variability. The resulting overall uncertainty on SWE is 15.6%. The proposed model has the potential to be a valuable tool to estimate the SWE from the only HS measurement in the Italian Alps, presenting even better performances during the late season (13.9%, 12.9% and 14.3% in March, April and May, respectively) that makes it particularly suitable for snow-hydrology studies.

Pursuit of Optimal Design for Winter-Balance Surveys of Valley-Glacier Ablation Areas

Efficient collection of snow depth and density data is important in field surveys used to estimate the winter surface mass balance of glaciers. Simultaneously extensive, high resolution and accurate snow-depth measurements can be difficult to obtain, so optimization of measurement configuration and spacing is valuable in any survey design. Using in-situ data from the ablation areas of three glaciers in the St. Elias Mountains of Yukon, Canada, we consider six possible survey designs for snow-depth sampling and $N$=6--200+ sampling locations per glacier. For each design and number of sampling locations, we use a linear regression on topographic parameters to estimate winter balance at unsampled locations and compare these estimates with known values. Average errors decrease sharply with increasing sample size up to $N\approx$10--15, but reliable error reduction for any given sampling scheme requires significantly higher $N$. Lower errors are often, but not always, associated with sampling schemes that employ quasi-regular spacing. With both real- and synthetic data, the common centreline survey produces the poorest results overall. The optimal design often requires sampling near the glacier margin, even at low $N$. The unconventional ``hourglass'' design performed best of all designs tested when evaluated against known values of winter balance.

Sub-kilometer Precipitation Datasets for Snowpack and Glacier Modeling in Alpine Terrain

apturing the spatial and temporal variability of precipitation at fine scale is necessary for high-resolution modeling of snowpack and glacier mass balance in alpine terrain. In this study, we assess the impact of three sub-kilometer precipitation datasets on distributed simulations of snowpack and glacier mass balance with the detailed snowpack model Crocus for winter 2011-2012. The different precipitation datasets at 500-m grid spacing over the northern and central French Alps are coming from (i) the SAFRAN reanalysis specially developed for alpine terrain interpolated at 500-m grid spacing, (ii) the numerical weather prediction (NWP) system AROME at 2.5-km resolution downscaled with a precipitation-elevation adjustment factor and (iii) a version of AROME at 500-m grid spacing. The spatial patterns of seasonal snowfall are first analyzed for the different precipitation datasets. Large differences between SAFRAN and the two versions of AROME are found at high-altitude and in regions of strong orographic precipitation enhancement. Results of Crocus snowpack simulations are then evaluated against (i) point measurements of snow depth, (ii) maps of snow covered areas retrieved from optical satellite data (MODIS) and (iii) field measurements of winter accumulation of six glaciers. The two versions of AROME lead to an overestimation of snow depth and snow-covered area, which are substantially improved by SAFRAN. However, all the precipitation datasets lead to an underestimation of snow depth increase at the daily scale and cumulated over the season, with AROME 500 m providing the best performances at the seasonal scale. The low correlation found between the biases in snow depth and in cumulated snow depth increase illustrates that total snow depth has a limited significance for the evaluation of precipitation datasets. Measurements of glacier winter mass balance showed a systematic underestimation of high-elevation snow accumulation with SAFRAN. The two versions of AROME overestimate the winter mass balance at four glaciers and produce nearly unbiased estimation for two of them. Our study illustrates the need for improvements in the precipitation field from high-resolution NWP systems for snow and glacier modeling in alpine terrain.

A novel algorithm for feature selection based on geographic distance metric: a case study of streamflow forecasting of Austria’s water resources

This paper focuses on input variable selection—feature selection—methods with the artificial neural network for the streamflow forecasting of large basins that have a variety of numerous stations. The feature selection methods in the current hydrology research community are not able to handle the problem in such basins. The paper proposes a novel feature selection algorithm—Bubble Selection—based on the idea of utilizing geographic distance as a metric. Evaluation of the performance of the algorithm is carried out by applying the Bubble Selection, to the case study of modeling Austria’s water resources of 540 stations in a single run mode. The aim is to select features for each station among 2412 stations, streamflow, precipitation, snow, snow depth, and water level measurements are available. The proposed algorithm allows considerably reducing the dimension of features. The Bubble Selection algorithm is further combined with the Sequential Forward Selection algorithm. Performance of the hybrid model is compared with the performance of Feature Ranking method in terms of the coefficient of determination, Nash–Sutcliffe Efficiency, and percent bias. The results show the superiority of the proposed hybrid algorithm over the Feature Ranking. The paper introduces a methodology to model a large basin and it reveals some skills that a feature selection algorithm should have.

Near Real-Time Measurement of Snow Water Equivalent in the Nepal Himalayas

Seasonal snow is an important component of the Himalayan hydrological system, but a lack of observations at high altitude hampers understanding and forecasting of water availability in this region. Here, we use a passive gamma ray sensor that measures snow water equivalent (SWE) and complementary meteorological instruments installed at 4962 m a.s.l. in the Nepal Himalayas to quantify the evolution of SWE and snow depth over a two year period. We assess the accuracy, spatial representativeness and the applicability of the SWE and snow depth measurements using time lapse camera imagery and field observations. The instrument setup performs well for snowpacks >50 mm SWE, but caution must be applied when interpreting measurements from discontinuous, patchy snow cover or those that contain lenses of refrozen meltwater. Over their typical ~6 month lifetime, snowpacks in this setting can attain up to 200 mm SWE, of which 10-15 % consists of mixed precipitation and rain-on-snow events. Precipitation gauges significantly underrepresent the solid fraction of precipitation received at this elevation by almost 40 % compared to the CS725. The application of sub-daily time lapse camera imagery can help to correctly interpret and increase the reliability and representativeness of snowfall measurements. Our monitoring approach provides high quality, continuous, near real time information that is essential to develop snow models in this data scarce region. We recommend that a similar instrument setup be extended into remote Himalayan environments to facilitate widespread snowpack monitoring and further our understanding of the high-altitude water cycle.

Converting snow depth to snow water equivalent using climatological variables

Abstract. We present a simple method that allows snow depth measurements to be converted to snow water equivalent (SWE) estimates. These estimates are useful to individuals interested in water resources, ecological function, and avalanche forecasting. They can also be assimilated into models to help improve predictions of total water volumes over large regions. The conversion of depth to SWE is particularly valuable since snow depth measurements are far more numerous than costlier and more complex SWE measurements. Our model regresses SWE against snow depth (h), day of water year (DOY) and climatological (30-year normal) values for winter (December, January, February) precipitation (PPTWT), and the difference (TD) between mean temperature of the warmest month and mean temperature of the coldest month, producing a power-law relationship. Relying on climatological normals rather than weather data for a given year allows our model to be applied at measurement sites lacking a weather station. Separate equations are obtained for the accumulation and the ablation phases of the snowpack. The model is validated against a large database of snow pillow measurements and yields a bias in SWE of less than 2 mm and a root-mean-squared error (RMSE) in SWE of less than 60 mm. The model is additionally validated against two completely independent sets of data: one from western North America and one from the northeastern United States. Finally, the results are compared with three other models for bulk density that have varying degrees of complexity and that were built in multiple geographic regions. The results show that the model described in this paper has the best performance for the validation data sets.

Comparing Aerial Lidar Observations With Terrestrial Lidar and Snow‐Probe Transects From NASA's 2017 SnowEx Campaign

AbstractNASA's 2017 SnowEx field campaign at Grand Mesa, CO, generated Airborne Laser Scans (ALS), Terrestrial Laser Scans (TLS), and snow‐probe transects, which allowed for a comparison between snow depth measurement techniques. At six locations, comparisons between gridded ALS and TLS observations, at 1‐m resolution, had a median snow depth difference of 5 cm, root‐mean‐square difference of 16 cm, mean‐absolute difference of 10 cm, and 3‐cm difference in standard deviation. ALS generally had greater but similar snow depth values to TLS, and results were not sensitive to the gridded cell size between 0.5 and 5 m. The greatest disagreements were where snow‐off TLS scans had shrubs and high incidence angles, leading to deeper snow depths (>10 cm) from ALS than TLS. The low vegetation and oblique angles caused occlusion in the TLS data and thus produced higher snow‐off bare Earth models relative to the ALS. Furthermore, in subcanopy areas where both ALS and TLS data existed, snow depth differences were comparable to differences in the open. Meanwhile, median values from 52 snow‐probe transects and near‐coincident ALS data had a mean difference of 6 cm, root‐mean‐square difference of 8 cm, mean‐absolute difference of 7 cm, and a mean difference in the standard deviation of 1 cm. Snow depth probes had greater but similar snow depth values to ALS. Therefore, based on comparisons with TLS and snow depth probes, ALS captured snow depth magnitude with better than or equal agreement to what has been reported in previous studies and showed the ability to capture high‐resolution spatial variability.

AMSR2 snow depth downscaling algorithm based on a multifactor approach over the Tibetan Plateau, China

Accurate dynamic measurements of snow depth (SD) on the Tibetan Plateau (TP) are of great importance for understanding local climate systems and surface hydrological cycles. Currently, passive microwave remote sensing is the most effective approach used to monitor the spatiotemporal variations in SD. However, the large uncertainties and low spatial resolution of existing SD products lead to less satisfactory passive microwave remote sensing in regions with complex terrain conditions, strong seasonal transitions and great spatiotemporal heterogeneity, such as the TP. In this study, geographic location, terrain, snow cover days and brightness temperature were evaluated. Then, critical factors were used to develop an Advanced Microwave Scanning Radiometer 2 (AMSR2) SD downscaling model. The results indicated the following: (1) SD is greatly influenced by geographic location, snow covered days, terrain parameters and brightness temperature difference. The surface roughness exhibited the best correlations with ground observations, which reflects 44% of the variation in SD; (2) the novel multifactor power SD downscaling model, which combines AMSR2 brightness temperature data and other auxiliary data, showed higher accuracy and stability and was closely correlated with the SD ground observations, and it reflected 80% of the SD variation; (3) compared with the AMSR2 ascending and descending orbital SD products released by the Japan Aerospace Exploration Agency (JAXA) and the SD datasets in China from the Environmental and Ecological Science Data Center for West China (WESTDC), the proposed downscaled SD datasets for the TP were greatly improved, and the root-mean-square error (RMSE) and average absolute error (MAE) were greatly reduced (2.00 cm and 0.25 cm, respectively); (4) the downscaled SD datasets for the TP showed good accuracy (RMSE = 0.58 cm) in areas with a SD of <3 cm; and (5) the worst monitoring accuracy of the downscaled SD product was for grasslands, with a RMSE of 2.07 cm. The best accuracy of the downscaling SD product was found for bare land cover conditions, with an RMSE of only 0.41 cm.

A compilation of multitemporal TLS snow depth distribution maps at the Weisssee snow research site (Kaunertal, Austria)

AbstractThe study presents a data set of the interannual and intraannual snow depth distribution recorded by terrestrial laser scanning (TLS) scans at Weisssee snow research site in Austria between November 2014 and May 2018. The data set comprises 23 snow‐on digital elevation models, one snow‐off digital elevation model, and the difference raster calculated between a snow‐off and snow‐on scans. The relative accuracy of the TLS scans was determined by measuring the distance between snow‐free planes from the snow‐on and snow‐off scans and shows mean values smaller than 0.03 m and standard deviations ranging between 0.02 and 0.1 m. The reliability of the snow depths derived from TLS was further assessed by comparing snow depths from snow probing, Global Navigation Satellite System measurements, and continuous snow depth measurements from the weather station. Comparison of the different measurement methods shows average deviations of less than 0.1 m. The data can be used for analysis and modeling of snow distribution or for assessing the representativeness of snow sensors or other remote sensing products.

Exploration of the Snow Ablation Process in the Semiarid Region in China by Combining Site-Based Measurements and the Utah Energy Balance Model—A Case Study of the Manas River Basin

Understanding the snow accumulation and melting process is of great significance for the assessment and regulation of water resources and the prevention of meltwater flooding, especially for the semiarid region in the Manas River Basin. However, the lack of long snow measurement time series in this semiarid region prevents a full understanding of the detailed local-scale snow ablation process. Additionally, the modeling of snow accumulation and melting is challenging due to parameter uncertainty. In this study, the snow ablation process in the Manas River Basin was quantitatively explored with long time-series of 3-h measurements of snow depth, snow density and snow water equivalent (SWE) at the Wulanwusu (WLWS), Hanqiazi (HQZ), and Baiyanggou (BYG) sites. This study explored the ability of the Utah energy balance (UEB) snow accumulation and melt model to simulate SWE, energy flux and water loss in the study area. Furthermore, the uncertainty in the ground surface aerodynamic roughness index zos in the UEB model was also analyzed. The results showed that: (1) noticeable variations in snow depth, SWE and snow density occurred on seasonal and interannual time scales, and variations in melting time and melting ratios occurred on short time scales; (2) a rapid decrease in snow depth did not influence the variations in SWE, and snow melting occurred during all time periods, even winter, which is a typical characteristic of snow accumulation in arid environments; (3) the UEB model accurately simulated the snow ablation processes, including SWE, snow surface temperature, and energy flux, at WLWS, HQZ, and BYG sites; (4) the lowest contribution of net radiation to melting occurred in the piedmont clinoplain, followed by the mountain desert grassland belt and mountain forest belt, whereas the contributions of net turbulence exhibited the opposite pattern; (5) the optimal zos in the UEB model was experimentally determined to be 0.01 m, and the UEB model-simulated SWE based on this value was the most consistent with the measured SWE; and (6) the results may provide theoretical and data foundations for research on the snow accumulation process at the watershed scale.

The Utility of Infrequent Snow Depth Images for Deriving Continuous Space‐Time Estimates of Seasonal Snow Water Equivalent

AbstractSnow water equivalent (SWE), particularly in mountains regions, has been an elusive hydrologic measurement. We examine the utility of a data assimilation approach to generate space‐time continuous estimates of SWE from more readily available snow depth (SD) measurements. A multitemporal lidar data set provides a unique opportunity to assimilate single SD images and verify posterior estimates against SD images at nonassimilation times. Application over three water years shows significant improvement in the posterior estimates with an average correlation between estimated and measured SD fields of 0.88 compared to 0.52 for prior estimates that do not benefit from the assimilated SD data. We also show that posterior estimates are consistent with independent in situ SWE and streamflow measurements. This work demonstrates that using high‐resolution/high‐accuracy, but infrequent, SD measurements combined with a data assimilation framework could make significant inroads toward the goal of spatially distributed SWE and snowmelt estimates at the global scale.

Improving Snow Water Equivalent Maps With Machine Learning of Snow Survey and Lidar Measurements

AbstractIn the semiarid interior western USA, where a majority of surface water supply comes from mountain forests, high‐resolution aerial lidar‐based surveys are commonly used to study snow. These surveys provide rich information about snow depth, but they are usually not accompanied with spatially explicit measurements of snow density, which leads to uncertainty in the estimation of snow water equivalent (SWE). In this study, we use a novel approach to distribute ~300 field measurements of snow density with artificial neural networks. We combine the resulting density maps with aerial lidar snow depth measurements, bias corrected with a very large and precisely geolocated array of field‐measured snow depths (~4,000 observations), to create and validate maps of snow depth, snow density, and SWE over two sites along Arizona's Mogollon Rim in February and March 2017. These maps show differences between midwinter and late‐winter snow conditions. In particular, compared to that of snow depth, the spatial variability of snow density is smaller for the later snow survey than the earlier snow survey. These gridded data also show that the representativeness of Snow Telemetry and other point measurements is different for the midwinter and late‐winter snow surveys. Overall, the lidar artificial neural network SWE estimates can be as much as 30% different than if Snow Telemetry density were used with lidar snow depths to estimate SWE.

Theia Snow collection: high-resolution operational snow cover maps from Sentinel-2 and Landsat-8 data

Abstract. The Theia Snow collection routinely provides high-resolution maps of the snow-covered area from Sentinel-2 and Landsat-8 observations. The collection covers selected areas worldwide, including the main mountain regions in western Europe (e.g. Alps, Pyrenees) and the High Atlas in Morocco. Each product of the Theia Snow collection contains four classes: snow, no snow, cloud and no data. We present the algorithm to generate the snow products and provide an evaluation of the accuracy of Sentinel-2 snow products using in situ snow depth measurements, higher-resolution snow maps and visual control. The results suggest that the snow is accurately detected in the Theia snow collection and that the snow detection is more accurate than the Sen2Cor outputs (ESA level 2 product). An issue that should be addressed in a future release is the occurrence of false snow detection in some large clouds. The snow maps are currently produced and freely distributed on average 5 d after the image acquisition as raster and vector files via the Theia portal (https://doi.org/10.24400/329360/F7Q52MNK).

Spatial controls on the distribution and dynamics of a marginal snowpack in the Australian Alps

AbstractSeasonal snowpacks in marginal snow environments are typically warm and nearly isothermal, exhibiting high inter‐ and intra‐annual variability. Measurements of snow depth and snow water equivalent were made across a small subalpine catchment in the Australian Alps over two snow seasons in order to investigate the extent and implications of snowpack spatial variability in this marginal setting. The distribution and dynamics of the snowpack were found to be influenced by upwind terrain, vegetation, solar radiation, and slope. The role of upwind vegetation was quantified using a novel parameter based on gridded vegetation height. The elevation range of the catchment was relatively modest (185 m), and elevation impacted distribution but not dynamics. Two characteristic features of marginal snowpack behaviour are presented. Firstly, the evolution of the snowpack is described in terms of a relatively unstable accumulation state and a highly stable ablation state, as revealed by temporal variations in the mean and standard deviation of snow water equivalent. Secondly, the validity of partitioning the snow season into distinct accumulation and ablation phases is shown to be compromised in such a setting. Snow at the most marginal locations may undergo complete melt several times during a season and, even where snow cover is more persistent, ablation processes begin to have an effect on the distribution of the snowpack early in the season. Our results are consistent with previous research showing that individual point measurements are unable to fully represent the variability in the snowpack across a catchment, and we show that recognising and addressing this variability are particularly important for studies in marginal snow environments.

Understanding subgrid variability of snow depth at 1‐km scale using Lidar measurements

AbstractIt is well known that snow plays an important role in land surface energy balance; however, modelling the subgrid variability of snow is still a challenge in large‐scale hydrological and land surface models. High‐resolution snow depth data and statistical methods can reveal some characteristics of the subgrid variability of snow depth, which can be useful in developing models for representing such subgrid variability. In this study, snow depth was measured by airborne Lidar at 0.5‐m resolution over two mountainous areas in south‐western Wyoming, Snowy Range and Laramie Range. To characterize subgrid snow depth spatial distribution, measured snow depth data of these two areas were meshed into 284 grids of 1‐km × 1‐km. Also, nine representative grids of 1‐km × 1‐km were selected for detailed analyses on the geostatistical structure and probability density function of snow depth. It was verified that land cover is one of the important factors controlling spatial variability of snow depth at the 1‐km scale. Probability density functions of snow depth tend to be Gaussian distributions in the forest areas. However, they are eventually skewed as non‐Gaussian distribution, largely due to the no‐snow areas effect, mainly caused by snow redistribution and snow melt. Our findings show the characteristics of subgrid variability of snow depth and clarify the potential factors that need to be considered in modelling subgrid variability of snow depth.

Retrieval of snow freeboard of Antarctic sea ice using waveform fitting of CryoSat-2 returns

Abstract. In this paper we develop a CryoSat-2 algorithm to retrieve the surface elevation of the air–snow interface over Antarctic sea ice. This algorithm utilizes a two-layer physical model that accounts for scattering from a snow layer atop sea ice as well as scattering from below the snow surface. The model produces waveforms that are fit to CryoSat-2 level 1B data through a bounded trust region least-squares fitting process. These fit waveforms are then used to track the air–snow interface and retrieve the surface elevation at each point along the CryoSat-2 ground track, from which the snow freeboard is computed. To validate this algorithm, we compare retrieved surface elevation measurements and snow surface radar return power levels with those from Operation IceBridge, which flew along a contemporaneous CryoSat-2 orbit in October 2011 and November 2012. Average elevation differences (standard deviations) along the flight lines (IceBridge Airborne Topographic Mapper, ATM – CryoSat-2) are found to be 0.016 cm (29.24 cm) in 2011 and 2.58 cm (26.65 cm) in 2012. The spatial distribution of monthly average pan-Antarctic snow freeboard found using this method is similar to what was observed from NASA's Ice, Cloud, and land Elevation Satellite (ICESat), where the difference (standard deviation) between October 2011–2017 CryoSat-2 mean snow freeboard and spring 2003–2007 mean freeboard from ICESat is 1.92 cm (9.23 cm). While our results suggest that this physical model and waveform fitting method can be used to retrieve snow freeboard from CryoSat-2, allowing for the potential to join laser and radar altimetry data records in the Antarctic, larger (∼30 cm) regional differences from ICESat and along-track differences from ATM do exist, suggesting the need for future improvements to the method. Snow–ice interface elevation retrieval is also explored as a potential to obtain snow depth measurements. However, it is found that this retrieval method often tracks a strong scattering layer within the snow layer instead of the actual snow–ice interface, leading to an overestimation of ice freeboard and an underestimation of snow depth in much of the Southern Ocean but with promising results in areas such as the East Antarctic sector.

Estimating Snow Depth and Leaf Area Index Based on UAV Digital Photogrammetry.

This study presents a novel approach in the application of Unmanned Aerial Vehicle (UAV) imaging for the conjoint assessment of the snow depth and winter leaf area index (LAI), a structural property of vegetation, affecting the snow accumulation and snowmelt. The snow depth estimation, based on a multi-temporal set of high-resolution digital surface models (DSMs) of snow-free and of snow-covered conditions, taken in a partially healthy to insect-induced Norway spruce forest and meadow coverage area within the Šumava National Park (Šumava NP) in the Czech Republic, was assessed over a winter season. The UAV-derived DSMs featured a resolution of 0.73–1.98 cm/pix. By subtracting the DSMs, the snow depth was determined and compared with manual snow probes taken at ground control point (GCP) positions, the root mean square error (RMSE) ranged between 0.08 m and 0.15 m. A comparative analysis of UAV-based snow depth with a denser network of arranged manual snow depth measurements yielded an RMSE between 0.16 m and 0.32 m. LAI assessment, crucial for correct interpretation of the snow depth distribution in forested areas, was based on downward-looking UAV images taken in the forest regime. To identify the canopy characteristics from downward-looking UAV images, the snow background was used instead of the sky fraction. Two conventional methods for the effective winter LAI retrieval, the LAI-2200 plant canopy analyzer, and digital hemispherical photography (DHP) were used as a reference. Apparent was the effect of canopy density and ground properties on the accuracy of DSMs assessment based on UAV imaging when compared to the field survey. The results of UAV-based LAI values provided estimates were comparable to values derived from the LAI-2200 plant canopy analyzer and DHP. Comparison with the conventional survey indicated that spring snow depth was overestimated, and spring LAI was underestimated by using UAV photogrammetry method. Since the snow depth and the LAI parameters are essential for snowpack studies, this combined method here will be of great value in the future to simplify snow depth and LAI assessment of snow dynamics.