Search

Monitoring properties of ice sheets in polar regions is one of the main challenges in glaciology. There is a large amount of heterogeneous radar data from the polar regions that have been gathered through expensive missions. However, retrieving meaningful information from this large volume of data is still a great challenge. With the advancement of machine learning techniques in recent years, many scientists are eager to take advantage of these algorithms and techniques to explore and mine Arctic and Antarctic data. These advancements, however, have happened mainly in the area of supervised learning where the models are data hungry and require large amounts of annotated data. Generating simulated data can be an effective and inexpensive approach to provide large labeled datasets for training machine learning models. In this work, we explore two approaches to simulate arctic snow radar echogram images, namely a radar scattering physics based approach combined with some statistical measures and a purely data-driven approach based on a conditional generative adversarial network. Using several image comparison metrics, we analyze the utility of both methods for the purpose of simulating echograms. Our results show that the physics simulator generates images with good structural similarities, while the purely data-driven approach achieves better textural similarities for simulated image. Finally, we also show that by augmenting our real dataset by the simulated echograms, we can improve our deep learning model for tracking internal layers of snow.

In this study, our goal is to track internal ice layers on the Snow Radar data collected by NASA Operation IceBridge. We examine the application of deep learning methods on radar data gathered from polar regions. Artificial intelligence techniques have displayed impressive success in many practical fields. Deep neural networks owe their success to the availability of massive labeled data. However, in many real-world problems, even when a large dataset is available, deep learning methods have shown less success, due to causes such as lack of a large labeled dataset, presence of noise in the data or missing data. In our radar data, the presence of noise is one of the main obstacles in utilizing popular deep learning methods such as transfer learning. Our experiments show that if the neural network is trained to detect contours of objects in electro-optical imagery, it can only track a low percentage of contours in radar data. Fine-tuning and further training do not provide any better results. However, we show that selecting the right model and training it on the radar imagery from the start yields far better results.

We present an ultrawideband (UWB) frequency-modulated (FM) microwave radar instrument capable of preserving cm-scale vertical resolution from a very long range (verified up to ~6 km). The system can operate either as a conventional continuous-wave FM radar (2–18 GHz) or in time-delayed stretch processing mode (2–14 GHz), providing high sensitivity, fine range resolution, and very low-range sidelobes. We present an overview of the system and laboratory test results to validate its performance. As a practical demonstration of its capabilities, we tested the system’s ability to map snow cover thickness on the Arctic marine ice. We present sample results from airborne measurements carried out at altitudes between ~500 m and ~6 km above the ground level onboard a long-range fixed-wing platform.

Radar sounding is a powerful geophysical approach for characterizing the subsurface conditions of terrestrial and planetary ice masses at local to global scales. As a result, a wide array of orbital, airborne, ground-based, and in situ instruments, platforms and data analysis approaches for radioglaciology have been developed, applied or proposed. Terrestrially, airborne radar sounding has been used in glaciology to observe ice thickness, basal topography and englacial layers for five decades. More recently, radar sounding data have also been exploited to estimate the extent and configuration of subglacial water, the geometry of subglacial bedforms and the subglacial and englacial thermal states of ice sheets. Planetary radar sounders have observed, or are planned to observe, the subsurfaces and near-surfaces of Mars, Earth's Moon, comets and the icy moons of Jupiter. In this review paper, and the thematic issue of theAnnals of Glaciologyon ‘Five decades of radioglaciology’ to which it belongs, we present recent advances in the fields of radar systems, missions, signal processing, data analysis, modeling and scientific interpretation. Our review presents progress in these fields since the last radio-glaciologicalAnnals of Glaciologyissue of 2014, the context of their history and future prospects.

Ice sheets reshape Earth’s surface. Maps of the landscape formed by past ice sheets are our best tool for reconstructing historic ice sheet behavior. But models of glacier erosion and deposition that explain mapped features are relatively untested, and without observations of landforms developing in situ, postglacial landscapes can provide only qualitative insight into past ice sheet conditions. Here we present the first swath radar data collected in Antarctica, demonstrating the ability of swath radar technology to map the subglacial environment of Thwaites Glacier (West Antarctica) at comparable resolutions to digital elevation models of deglaciated terrain. Incompatibility between measured bedform orientation and predicted subglacial water pathways indicates that ice, not water, is the primary actor in initiating bedform development at Thwaites Glacier. These data show no clear relationship between morphology and glacier speed, a weak relationship between morphology and basal shear stress, and highlight a likely role for preexisting geology in glacial bedform shape.

We developed a high-performance, multichannel, ultra-wideband radar system for measurements of the base and interior of the East Antarctic Ice Sheet. We designed the radar to be of high power (4000-W peak) yet portable and to be able to operate with 60-MHz bandwidth at a center frequency of 200 MHz, providing high sensitivity and fine vertical resolution relative to current technology. We used the radar to perform extensive mea- surements as a part of a multinational collaboration. We collected data onboard a tracked vehicle outfitted with an array of high-gain antennas. We sounded 2- to 3-km thick ice near Dome Fuji. Prelim- inary ice thickness data match those obtained via semicoincident measurements performed with a different surface-based pulse- modulated radar system operated during the same field campaign, as well as previous airborne measurements. In addition, we mapped internal reflection horizons with fine vertical resolution from 300 m below the ice surface to ∼100 m above the bed. In this article, we provide a detailed overview of the radar instrument design, implementation, and field measurement setup. We present sample data to illustrate its capabilities and discuss how the data collected with it will be valuable for the assessment of promising drilling sites to recover ice cores that are 0.9–1.5 million years old.

We present a high-resolution airborne radar data set (EGRIP-NOR-2018) for the onset region of the Northeast Greenland Ice Stream (NEGIS). The radar data were acquired in May 2018 with Alfred Wegener Institute’s multichannel ultra-wideband (UWB) radar mounted on the Polar6 aircraft. Radar profiles cover an area of ~24000 km2 and extend over the well-defined shear margins of the NEGIS. The survey area is centred at the location of the drill site of the East Greenland Ice-Core Project (EastGRIP) and several radar lines intersect at this location. The survey layout was designed to: (i) map the stratigraphic signature of the shear margins with radar profiles aligned perpendicular to ice flow, (ii) trace the radar stratigraphy along several flow lines and (iii) provide spatial coverage of ice thickness and basal properties. While we are able to resolve radar reflections in the deep stratigraphy, we can not fully resolve the steeply inclined reflections at the tightly folded shear margins in the lower part of the ice column. The NEGIS is causing the most significant discrepancies between numerically modelled and observed ice surface velocities. Given the high likelihood of future climate and ocean warming, this extensive data set of new high-resolution radar data in combination with the EastGRIP ice core will be a key contribution to understand the past and future dynamics of the NEGIS. The EGRIP-NOR-2018 radar data products can be obtained at the PANGAEA Data Publisher (https://doi.pangaea.de/10.1594/PANGAEA.928569; Franke et al. 2021a).

Tree heights are important input for many inventory and ecosystem models. While optical and infrared sensors such as LiDAR are widely used in forest surveys, microwave radar has the unique advantage of being able to operate in bad weather or poor visibility conditions. In this work, we employed a LiDAR combined with a compact ultra-wideband frequency modulated continuous wave (FMCW) radar onboard a Single Otter aircraft to collect data over forested areas in Alaska. While the primary focus of the mission was to map the surface elevation and snow thickness of Alaskan glaciers as a part of NASA Operation IceBridge (OIB), we recorded LiDAR and radar returns during the transit flights to analyze backscattering signatures from tree-covered areas, thereby assessing the potential application of our radar to forest and vegetation remote sensing. We analyzed these measurements and estimated the tree heights along the flight trajectory. The very good agreement between the tree height profiles from the two instruments shows promising results for wide area forestry studies using microwave radar.

This paper provides an update and overview of the Center for Remote Sensing of Ice Sheets (CReSIS) radars and platforms, including representative results from these systems. CReSIS radar systems operate over a frequency range of 14–38 GHz. Each radar system's specific frequency band is driven by the required depth of signal penetration, measurement resolution, allocated frequency spectra, and antenna operating frequencies (often influenced by aircraft integration). We also highlight recent system advancements and future work, including (1) increasing system bandwidth; (2) miniaturizing radar hardware; and (3) increasing sensitivity. For platform development, we are developing smaller, easier to operate and less expensive unmanned aerial systems. Next-generation platforms will further expand accessibility to scientists with vertical takeoff and landing capabilities.

Multichannel radar depth sounding systems are able to produce two-dimensional (2D) and three-dimensional (3D) imagery of the internal structure of polar ice sheets. Information such as ice thickness and surface elevation is extracted from these data and applied to research in ice flow modeling and ice mass balance calculations. Due to a large amount of data collected, we seek to automate the ice-bottom layer tracking and allow for efficient manual corrections when errors occur in the automated method. We present improvements made to previous implementations of the Viterbi and sequential tree-reweighted message passing (TRW-S) algorithms for ice-bottom extraction in 2D and 3D radar imagery. These improvements are in the form of novel cost functions that allow for the integration of further domain-specific knowledge into the cost calculations and provide additional evidence of the characteristics of the ice sheets surveyed. Along with an explanation of our modifications, we demonstrate the results obtained by our modified implementations of the two algorithms and by previously proposed solutions to this problem, when compared to manually corrected ground truth data. Furthermore, we perform a self-assessment of tracking results by analyzing differences in the estimated ice-bottom for surveyed locations where flight paths have crossed and, thus, two separate measurements have been made at the same location. Using our modified cost functions and preprocessing routines, we obtain significantly decreased mean error measurements from both algorithms, such as a 47% reduction in average tracking error in the case of 3D imagery between the original and our proposed implementation of TRW-S.

Ice thickness estimation is an important aspect of ice sheet studies. In this work, we use convolutional neural networks with multiple output nodes to regress and learn the thickness of internal ice layers in Snow Radar images collected in northwest Greenland. We experiment with some state-of-the-art networks and find that with the residual connections of ResNet50, we could achieve a mean absolute error of 1.251 pixels over the test set. Such regression-based networks can further be improved by embedding domain knowledge and radar information in the neural network in order to reduce the requirement of manual annotations.

The North East Greenland Ice Stream clearly stands out in the surface velocity field of the ice flow of Greenland, with its sharp and narrow shear margins visible in the flow field almost up to the central divide. While the current extent and strength of the streaming can be determined from remotely sensed velocities of the ice surface, it is less known how the ice stream is affecting the deeper layers of ice in its catchment area, and how it may have evolved over time. The deformation of the ice due to streaming can be made visible by mapping the distortion of the isochronous stratigraphy of the ice. This has been done by an airborne radar survey centering on the location of the EGRIP drilling camp, carried out with the ultra wide band radar system (AWI UWB). The dense grid of profiles arranged mainly perpendicular to the ice flow reveals the imprint that the strong shearing leaves within the layering of the ice. Although the layers are tightly folded and distorted within the shear zones, it is possible to continuously trace reflections within the upper half of the ice column throughout the entire survey area. It can be shown that the intensity of the folding is linked to the strain rate field derived from the surface velocities, and that the deformation history of the ice is preserved in the folded layers, even after it is no longer affected by high strain rates. The advection patterns of the mapped stratigraphic features reveal how the streaming of the ice and the resulting local changes of surface topography may have affected the inflow into the stream and the position of the shear margins over time.

Crystal anisotropy of ice causes slight birefringence for electromagnetic waves. At the same time, the mechanical anisotropy amounts to several orders of magnitude, thus making fabric properties highly-relevant for internal deformation. To date, bulk anisotropy of glaciers and ice sheets can be determined by geophysical methods, such as polarimetric radar, or direct sampling from ice cores. A shortcoming has been so far that changes of bulk anisotropy could mainly be inferred at single point observations, but less so as continuous profiles. Here, we exploit the effect of birefringence caused by bulk anisotropy in co-polarized airborne radar data to determine the horizontal anisotropy across the North-East Greenland Ice Stream. We base our analysis on the fact that birefringence causes a second-order effect on radar amplitudes, which leads to a beat frequency in the low and medium frequency range (O(100 kHz)), which is proportional to the horizontal anisotropy. Complementing our radar analysis with direct fabric and dielectric property observations we can constrain the range of all three fabric eigenvalues as a function of space across and along the ice stream. Finally, we assess the effect of the inferred fabric distribution on the overall ice rheology in the context of ice stream dynamics and compare it with numerical model results. Our overall approach has the advantage that it can be applied to co-polarized radar systems, as commonly used in profiling surveys, and does not require dedicated cross-polarized radar set-up. This provides the opportunity to revisit older data, especially from Greenland and Antarctica, to map fabric anisotropy in ice-dynamically interesting regions.

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 in situ 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.

We present an ice-penetrating radar data set which consists of 26 internal reflecting horizons (IRHs) that cover the entire Dome C area of the East Antarctic plateau, the most extensive to date in the region. This data set uses radar surveys collected over the space of 10 years, starting with an airborne international collaboration in 2008 to explore the region, up to the detailed ground based surveys in support of the Beyond EPICA – Oldest Ice (BE-OI) European Consortium. Through direct correlation with the EPICA-DC ice core, we date 19 IRHs that span the past four glacial cycles, from the beginning of the Holocene to over 350 ka. We indirectly date and provide stratigraphic information for seven older IRHs using an 1-D ice flow inverse model, going back to an estimated 700 ka. Depth and age uncertainties are quantified for all IRHs and provided as part of the data set. The IRH data set presented in this study is available at the U.S. Antarctic Program Data Center (USAP-DC) (https://doi.org/10.15784/601411, Cavitte et al., 2020) and represents a contribution to the SCAR AntArchitecture program.

Global warming is rapidly reducing glaciers and ice sheets across the world. Real time assessment of this reduction is required so as to monitor its global climatic impact. In this paper, we introduce a novel way of estimating the thickness of each internal ice layer using Snow Radar images and Fully Convolutional Networks. The estimated thickness can be used to understand snow accumulation each year. To understand the depth and structure of each internal ice layer, we perform multi-class semantic segmentation on radar images, which hasn't been performed before. As the radar images lack good training labels, we carry out a pre-processing technique to get a clean set of labels. After detecting each ice layer uniquely, we calculate its thickness and compare it with the processed ground truth. This is the first time that each ice layer is detected separately and its thickness calculated through automated techniques. Through this procedure we were able to estimate the ice-layer thicknesses within a Mean Absolute Error of approximately 3.6 pixels. Such a Deep Learning based method can be used with ever-increasing datasets to make accurate assessments for cryospheric studies.

Airborne sounders with cross-track antenna arrays coherently combine measurements from multiple sensors to isolate nadir echoes from co-range lateral surface clutter. The array size, which is limited by the constraints of the platform, determines the degree to which clutter may be suppressed. Array processing techniques offer strategic advantages for suppressing interference in the cross-track dimension but require a well constrained model of the array's response to directional sources (referred to as the array manifold). We propose and test an empirical array manifold characterization using a fine-resolution digital elevation model of the Arctic and multibeam data collected by a Center for Remote Sensing of Ice Sheets (CReSIS) sensor, the Multichannel Coherent Radar Depth Sounder (MCoRDS), during NASA's Operation Ice-Bridge. The results outline our first efforts to estimate the manifold and evidence improvements observed in combined images.

We designed a Ka-band prototype radar altimeter operated at a center frequency of 35 GHz with a 6 GHz bandwidth. The instrument was intended for fine-resolution verification and validation of space borne altimetry datasets. We installed it onboard the NASA C-130 aircraft in conjunction with two other wideband microwave instruments and collected airborne altimetry data over Greenland land ice and arctic sea ice during the 2015 NASA Operation IceBridge arctic campaign. Apart from the major application of verification and calibration of satellite-based measurements, data from this instrument can be used to derive snow grain size because of the dominant effect of volume scattering in radar signatures. In this paper, we briefly describe the system design and the installation on the NASA C-130, discuss the observed penetration depths of Ka-band signals into the snowpack, present sample results of optical-equivalent snow grain size estimates from radar measurements over the dry snow zone using a simplified snowpack model. We show that the observed penetration depths and the snow grain size estimates from the airborne Ka-band radar retrievals agree well with the model and in-situ snow-pit measurements.

This paper presents preliminary results using a fully connected neural network (NN) to automatically track the internal layers of snow radar echograms using an iterative “row-block-column” approach. Snow radar images, when accurately tracked, provide relevant information for estimating snow accumulation rates in polar regions which is a key measurement needed to understand and predict the impact of climate warming in Greenland and Antarctica. A multiclass NN was designed and trained with a training set of 121,408 columns of simulated snow radar data and learns to automatically track the internal layers with an accuracy of 92.8%, a RMSE of 0.24 pixels, and with 98% of pixel errors less than or equal to 1 pixel.

Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. In this research, we evaluated the performance of synthetically generated snow radar images based on modified cycle-consistent adversarial networks. We conducted several experiments to test the quality of the generated radar imagery. Our experiments show a very good similarity between real and synthetic snow radar images.

Differential Interferometric Synthetic Aperture Radar (DIn-SAR) processing techniques applied to ice penetrating radar enable precise measurement of the vertical displacement of englacial layers within an ice sheet. This technique has primarily been applied using ground based ice-penetrating radar due to the ability to achieve a near-zero spatial baseline. We investigate this technique on data from the Multichannel Coherent Radar Depth Sounder (MCoRDS), an airborne ice penetrating radar, and produce initial results from a high accumulation region near Camp Century in northwest Greenland. We estimate the vertical displacement by compensating for the spatial baseline using precise trajectory information and estimates of the cross-track layer slope from direction of arrival analysis. The measurement accuracy is still being investigated.

The Northeast Greenland Ice Stream (NEGIS) is an important dynamic component for the total mass balance of the Greenland ice sheet, as it reaches up to the central divide and drains 12% of the ice sheet. The geometric boundary conditions and in particular the nature of the subglacial bed of the NEGIS are essential to understand its ice flow dynamics. We present a record of more than 8000 km of radar survey lines of multi-channel, ultra-wideband radio echo sounding data covering an area of 24 000 km2, centered on the drill site for the East Greenland Ice-core Project (EGRIP), in the upper part of the NEGIS catchment. Our data yield a new detailed model of ice-thickness distribution and basal topography in the region. The enhanced resolution of our bed topography model shows features which we interpret to be caused by erosional activity, potentially over several glacial–interglacial cycles. Off-nadir reflections from the ice–bed interface in the center of the ice stream indicate a streamlined bed with elongated subglacial landforms. Our new bed topography model will help to improve the basal boundary conditions of NEGIS prescribed for ice flow models and thus foster an improved understanding of the ice-dynamic setting.

Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. In this way, training data can be quickly augmented with additional images. In this research, we evaluated the performance of synthetically generated radar images based on modified cycle-consistent adversarial networks. We conducted several experiments to test the quality of the generated radar imagery. We also tested the quality of a state-of-the-art contour detection algorithm on synthetic data and different combinations of real and synthetic data. Our experiments show that synthetic radar images generated by generative adversarial network (GAN) can be used in combination with real images for data augmentation and training of deep neural networks. However, the synthetic images generated by GANs cannot be used solely for training a neural network (training on synthetic and testing on real) as they cannot simulate all of the radar characteristics such as noise or Doppler effects. To the best of our knowledge, this is the first work in creating radar sounder imagery based on generative adversarial network.

Artificial intelligence (AI) techniques have displayed impressive success in many practical fields. Deep neural networks (DNNs) owe their success to the availability of massive labeled data. However, in many real-world problems, even when a large dataset is available, deep learning methods have shown less success, due to causes such as lack of large labeled dataset, presence of noise in data, or missing data. In the present work, we intend to examine the application of deep learning methods on radar data gathered from polar regions. Our goal is to track internal ice layers in radar imagery. In such data, the presence of noise is one of the main obstacles in utilizing popular deep learning methods such as transfer learning. Our experiments show that if the neural network is trained to detect contours of objects in electro-optical imagery, it can only track a low percentage of contours in radar data. Fine-tuning and further training do not provide any better results. However, we will show that selecting the right model and training the model on the radar imagery from the base, is going to yield far better results. We also discuss another possible learning approach that can save us time for data annotation.

The catchments of Pine Island Glacier and Thwaites Glacier in the Amundsen Sea Embayment are two of the largest, most rapidly changing, and potentially unstable sectors of the West Antarctic Ice Sheet. They are also neighboring outlets, separated by the topographically unconfined eastern shear margin of Thwaites Glacier and the southwest tributary of Pine Island Glacier. This tributary begins just downstream of the eastern shear margin and flows into the Pine Island ice shelf. As a result, it is a potential locus of interaction between the two glaciers and could result in cross-catchment feedback during the retreat of either. Here, we analyze relative basal reflectivity profiles from three radar sounding survey lines collected using the UTIG HiCARS radar system in 2004 and CReSIS MCoRDS radar system in 2012 and 2014 to investigate the extent and character of ocean access beneath the southwest tributary. These profiles provide evidence of ocean access ~12 km inland of the 1992–2011 InSAR-derived grounding line by 2014, suggesting either retreat since 2011 or the intrusion of ocean water kilometers inland of the grounding line.

Climate change is extensively affecting ice sheets resulting in accelerating mass loss in recent decades. Assessment of this reduction and its causes is required to project future ice mass loss. Annual snow accumulation is an important component of the surface mass balance of ice sheets. While in situ snow accumulation measurements are temporally and spatially limited due to their high cost, airborne radar sounders can achieve ice sheet wide coverage by capturing and tracking annual snow layers in the radar images or echograms. In this paper, we use deep learning to uniquely identify the position of each annual snow layer in the Snow Radar echograms taken across different regions over the Greenland ice sheet. We train with more than 15,000 images generated from radar echograms and estimate the thickness of each snow layer within a mean absolute error of 0.54 to 7.28 pixels, depending on dataset. A highly precise snow layer thickness can help improve weather models and, thus, support glaciological studies. Such a well-trained deep learning model can be used with ever-growing datasets to aid in the accurate assessment of snow accumulation on the dynamically changing ice sheets.

Since 2009, the ultra-wideband snow radar on Operation IceBridge (OIB; a NASA airborne mission to survey the polar ice covers) has acquired data in annual campaigns conducted during the Arctic and Antarctic springs. Progressive improvements in radar hardware and data processing methodologies have led to improved data quality for subsequent retrieval of snow depth. Existing retrieval algorithms differ in the way the air–snow (a–s) and snow–ice (s–i) interfaces are detected and localized in the radar returns and in how the system limitations are addressed (e.g., noise, resolution). In 2014, the Snow Thickness On Sea Ice Working Group (STOSIWG) was formed and tasked with investigating how radar data quality affects snow depth retrievals and how retrievals from the various algorithms differ. The goal is to understand the limitations of the estimates and to produce a well-documented, long-term record that can be used for understanding broader changes in the Arctic climate system. Here, we assess five retrieval algorithms by comparisons with field measurements from two ground-based campaigns, including the BRomine, Ozone, and Mercury EXperiment (BROMEX) at Barrow, Alaska; a field program by Environment and Climate Change Canada at Eureka, Nunavut; and available climatology and snowfall from ERA-Interim reanalysis. The aim is to examine available algorithms and to use the assessment results to inform the development of future approaches. We present results from these assessments and highlight key considerations for the production of a long-term, calibrated geophysical record of springtime snow thickness over Arctic sea ice.

Frequency-modulated-continuous-wave (FM-CW) radar has been used extensively for the airborne measurement of snow thickness over sea ice and the mapping of annual accumulation over land ice. In contrast to conventional in situ measurements, FM-CW radar, when operated onboard an airborne platform, can be a useful tool for widearea surveys of snow deposition. Since the early 2000s, the Center for Remote Sensing of Ice Sheets (CReSIS) at the University of Kansas (KU) has designed, developed, and deployed airborne ultrawide-band (UWB) FM-CW radars, called Snow Radars, on National Science Foundation (NSF)-, NASA-, Naval Research Laboratory (NRL)-, and Alfred Wegener Institute (AWI)-provided aircraft in both Arctic and Antarctic regions and generated a large amount of snow data products. In addition to the significant standalone value of the snow-thickness measurements, these data are being used in estimating Arctic sea ice thickness, which is a key variable in the study of atmosphere-ocean-ice interactions. This article provides a review of snow remote sensing techniques with airborne FM-CW radars to document the operating principle, design, and evolution of CReSIS' UWB FM-CW radars and discuss and promote understanding of the advantages and limitations associated with these systems.

Subglacial roughness can be determined at a variety of length scales from radio-echo sounding (RES) data either via statistical analysis of topography or inferred from basal radar scattering. Past studies have demonstrated that subglacial terrain exhibits self-affine (power law) roughness scaling behaviour, but existing radar scattering models do not take this into account. Here, using RES data from northern Greenland, we introduce a self-affine statistical framework that enables a consistent integration of topographic-scale roughness with the electromagnetic theory of radar scattering. We demonstrate that the degree of radar scattering, quantified using the waveform abruptness (pulse peakiness), is topographically controlled by the Hurst (roughness power law) exponent. Notably, specular bed reflections are associated with a lower Hurst exponent, with diffuse scattering associated with a higher Hurst exponent. Abrupt waveforms (specular reflections) have previously been used as a RES diagnostic for basal water, and to test this assumption we compare our radar scattering map with a recent prediction for the basal thermal state. We demonstrate that the majority of thawed regions (above pressure melting point) exhibit a diffuse scattering signature, which is in contradiction to the prior approach. Self-affine statistics provide a generalised model for subglacial terrain and can improve our understanding of the relationship between basal properties and ice-sheet dynamics.

Jakobshavn Isbrae, West Greenland, which holds a 0.6 m sea level volume equivalent, has been speeding up and retreating since the late 1990s. Interpretation of its retreat has been hindered by difficulties in measuring its ice thickness with airborne radar depth sounders. Here we employ high-resolution, helicopter-borne gravity data from 2012 to reconstruct its bed elevation within 50 km of the ocean margin using a three-dimensional inversion constrained by fjord bathymetry data offshore and a mass conservation algorithm inland. We find the glacier trough to be asymmetric and several 100 m deeper than estimated previously in the lower part. From 1996 to 2016, the grounding line migrated at 0.6 km/yr from 700 m to 1100 m depth. Upstream, the bed drops to 1600 m over 10 km then slowly climbs to 1200 m depth in 40 km. Jakobshavn Isbrae will continue to retreat along a retrograde slope for decades to come.

We present a compact, multi-channel, wideband VHF radar system for fine-resolution measurements of the base and interior of large ice sheets. Data from this radar will be used in the identification of potential drill locations to retrieve ice core samples more than 1 million years old in East Antarctica. The radar is a lightweight instrument equipped with four >1-kW peak power transmit channels, eight independent digital receivers, and an array of high-gain antennas. We developed the system as part of a collaborative effort between the United States of America, Japan and Norway. The instrument was used for surface-based surveys near Dome-Fuji onboard a tracked vehicle during the 2018/2019 Austral Summer, covering 2,000 line-km. This paper presents an overview of the radar system, accompanied by laboratory and field test results that demonstrate the system’s ability to map the internal ice sheet structure and basal conditions with outstanding detail.

Snow in mountainous areas provides freshwater resources for lower basins and coastal areas. Snow layering at mountain summits contains information about local seasonal snow accumulation and climate history. Knowledge of snow depths is required to estimate the snow water equivalent. However, monitoring spatial distribution and temporal changes in snow depth and accumulation over remote mountains and glaciers in wide areas is challenging because of limited accessibility for in-situ measurements. As a part of NASA Operation IceBridge missions, we took airborne radar measurements of snow over Alaskan mountains, icefields and glaciers in late May of 2018, with a compact frequency-modulated continuous wave radar system, operating from 2 GHz to 8 GHz and installed on a Single Otter aircraft. In this paper, we describe the radar instrument, its installation onto the platform, the data collection and processing activities, and report the major results from these surveys. We observed seasonal snow depth between ~0.3 m to ~15 m for elevations above sea level from ~1726 m to ~3624 m. We successfully mapped snow accumulation layers to depths exceeding ~85 m below the surface at high-elevation summits of Mount Wrangell and Bona. The traced snow depth profiles over glaciers and accumulation layers at mountain summits point to the utility of these data to the study of water resource management, hydrology modelling, and regional climate change.

The mapping of internal annual layers in polar firn with ultra-wideband (UWB) surface-based and airborne radars enables the assessment of past snow accumulation amounts and patterns. We used a UWB microwave radar (snow radar) to image and track internal layers in firn over hundreds of kilometers in West Antarctica (WA). We used dated ice-core records to determine that stratigraphic internal layers (SILs) are mainly formed in spring. The analysis uncertainty is generally small, with the highest values associated with regions and/or years with low accumulation. In survey regions with multiple ice-core records, the choice of ice-core record used for the radar analysis was found to have a very small contribution to the overall analysis uncertainty.

We developed a multichannel wideband synthetic aperture radar (SAR) that operates over a frequency range of 190–450 MHz for measurements over the ice sheets in Antarctica and Greenland. The antenna-array, which consists of eight elements housed in a certified external structure for a BASLER aircraft, was used for measurements during the 2013–2014 Antarctic field season. We performed measurements with this system in conjunction with two ultra-wideband radars operating over a frequency range of 2–8 GHz and 12–18 GHz on Siple Coast ice streams in West Antarctica during December 2013 and January 2014. We sounded ice thicker than 2 km with a signal-to-noise ratio (SNR) of more than 20 dB in an area with two-way ice loss of about 27 dB/km. The same system also simultaneously mapped near-surface internal layers with submeter resolution from the ice-surface to a depth of about 1100 for 1200 m thick ice. In this paper, we provide a detailed overview of the radar instrumentation and signal processing algorithms and present a few sample results. The radar will be operated over a frequency range of 150–550 MHz with a 24-element antenna-array for wide-ranging measurements over the Greenland and Antarctic ice sheets, starting around August 2015.

This paper presents recent updates to the CReSIS radar sensor package and the platforms supporting these sensors. These sensors cover a wide frequency range (14 MHz to 38 GHz). The specific frequency bands are chosen to balance between bandwidth available and signal penetration. The wide frequency range is also used for measuring different phenomenology. CReSIS has integrated these radar systems, including antennas, on a wide variety of fixed wing crewed aircraft, several UAV platforms, and for ground-based applications. The software for processing the radar data is now open source and an overview of the capabilities and how to access and use the software are presented. Finally, example data products which explore the new capabilities of the sensors and platforms are given.

The performance of several methods to estimate the number of source signals impinging on a sensor array are compared using a traditional simulator and their performance for synthetic aperture radar tomography is discussed. All methods use a penalty term that increases with model order in order to prevent overestimation. We include both separate estimation of model selection and direction of arrival as well as joint estimation. We formulate a new penalty term, numerically tuned so that it gives optimal performance over our operating conditions, and compare this method as well. Simulation results show that the numerically tuned model selection criteria is optimal and that the typical methods do not do well for low snapshots. We also found that there is little sensitivity to SNR greater than 3 dB when the number of snapshots is high. We discuss some issues to applying the algorithms to data collected by the CReSIS radar depth sounder.

Deep learning methods have surpassed the performance of traditional techniques on a wide range of problems in computer vision, but nearly all of this work has studied consumer photos, where precisely correct output is often not critical. It is less clear how well these techniques may apply on structured prediction problems where fine-grained output with high precision is required, such as in scientific imaging domains. Here we consider the problem of segmenting echogram radar data collected from the polar ice sheets, which is challenging because segmentation boundaries are often very weak and there is a high degree of noise. We propose a multi-task spatiotemporal neural network that combines 3D ConvNets and Recurrent Neural Networks (RNNs) to estimate ice surface boundaries from sequences of tomographic radar images. We show that our model outperforms the state-of-the-art on this problem by (1) avoiding the need for hand-tuned parameters, (2) extracting multiple surfaces (ice-air and ice-bed) simultaneously, (3) requiring less non-visual metadata, and (4) being about 6 times faster., 10 pages, 7 figures, published in WACV 2018

Ice properties inferred from multi-polarization measurements, such as birefringence and crystal orientation fabric (COF), can provide insight into ice strain, viscosity, and ice flow. In 2008, the Center for Remote Sensing of Ice Sheets (CReSIS) used a ground-based VHF (very high frequency) radar to take multi-channel and multi-polarization measurements around the NEEM (North Greenland Eemian Ice Drilling) site. The system operated with 30 MHz bandwidth at a center frequency of 150 MHz. This paper describes the radar system, antenna configurations, data collection, and processing and analysis of this data set. Within the framework derived from uniaxial ice crystal model, we found that ice birefringence dominates the power variation patterns of co-polarization and cross-polarization measurements in the area of 100 km2 around the ice core site. The phase shift between ordinary and extraordinary waves increases nonlinearly with depth. The ice optic axis lies in planes that are close to the vertical plane and perpendicular or parallel to the ice divide depending on depth. The ice optic axis has an average tilt angle of about 11.6° vertically, and its plane may rotate either clockwise or counterclockwise by about 10° across the 100 km2 area, and at a specific location the plane may rotate slightly counterclockwise as depth increases. Comparisons between the radar observations, simulations, and ice core fabric data are in very good agreement. We calculated the effective colatitude at different depths by using azimuth and colatitude measurements of the c axis of ice crystals. We obtained an average effective c axis tilt angle of 9.6° from the vertical axis, very comparable to the average optic axis tilt angle estimated from radar polarization measurements. The comparisons give us confidence in applying this polarimetric radio echo sounding technique to infer profiles of ice fabric in locations where there are no ice core measurements.

Ground-penetrating radar on planes and satellites now makes it practical to collect 3D observations of the subsurface structure of the polar ice sheets, providing crucial data for understanding and tracking global climate change. But converting these noisy readings into useful observations is generally done by hand, which is impractical at a continental scale. In this paper, we propose a computer vision-based technique for extracting 3D ice-bottom surfaces by viewing the task as an inference problem on a probabilistic graphical model. We first generate a seed surface subject to a set of constraints, and then incorporate additional sources of evidence to refine it via discrete energy minimization. We evaluate the performance of the tracking algorithm on 7 topographic sequences (each with over 3000 radar images) collected from the Canadian Arctic Archipelago with respect to human-labeled ground truth., 5 pages, 3 figures, published in ICIP 2017

The flow of ice is temperature-dependent, but direct measurements of englacial temperature are sparse. The dielectric attenuation of radio waves through ice is also temperature-dependent, and radar sounding of ice sheets is sensitive to this attenuation. Here we estimate depth-averaged radar-attenuation rates within the Greenland Ice Sheet from airborne radar-sounding data and its associated radiostratigraphy. Using existing empirical relationships between temperature, chemistry, and radar attenuation, we then infer the depth-averaged englacial temperature. The dated radiostratigraphy permits a correction for the confounding effect of spatially varying ice chemistry. Where radar transects intersect boreholes, radar-inferred temperature is consistently higher than that measured directly. We attribute this discrepancy to the poorly recognized frequency dependence of the radar-attenuation rate and correct for this effect empirically, resulting in a robust relationship between radar-inferred and borehole-measured depth-averaged temperature. Radar-inferred englacial temperature is often lower than modern surface temperature and that of a steady state ice-sheet model, particularly in southern Greenland. This pattern suggests that past changes in surface boundary conditions (temperature and accumulation rate) affect the ice sheet's present temperature structure over a much larger area than previously recognized. This radar-inferred temperature structure provides a new constraint for thermomechanical models of the Greenland Ice Sheet.

Satellite radar altimeter (RA) measurements are important for continued monitoring of rapidly changing polar regions. In 2010, the European Space Agency launched CryoSat-2 carrying SIRAL, a Ku-band RA with objectives of determining the thickness and extent of sea ice and the topography of the ice sheets. One difficulty with Ku-band radar surveys over snow and ice is unknown penetration of RA signal into snow cover. Improving our understanding of the interactions of RA signals with snow and ice is needed to produce accurate elevation products. To this end, we developed a low-power, ultrawideband (12–18 GHz) RA for airborne surveys to provide fine resolution measurements capable of detecting both scattering from the surface and layers within sea ice and ice sheets. These measurements provide a means of identifying the dominant scattering location of lower resolution RA measurements comparable to satellite-based instruments. We generated two products: a full-bandwidth waveform (FBW) to identify scattering targets at fine resolution and a reduced-bandwidth waveform (RBW) to represent conventional RA measurements. Retrackers are used to generate height estimates over various surface conditions for comparisons. Over ice sheets, the leading-edge tracker provided consistent ice-surface elevation measurements between the FBW and RBW results; however, there were significant differences between the results from the centroid tracker. Over sea ice, the location of the dominant return between the results from snow-covered sea ice is highly variable. This paper provides an overview of RA surveys in polar regions, a description of the CReSIS system, and a discussion of the results.

We have built and operated an ultra-wideband UHF pulsed-chirp radar for measuring firn stratigraphy from airborne platforms over the ice sheets of Greenland and West Antarctica. Our analysis found a wide range of capabilities, including imaging of post firn–ice transition horizons and sounding of shallow glaciers and ice shelves. Imaging of horizons to depths exceeding 600 m was possible in the colder interior regions of the ice sheet, where scattering from the ice surface and inclusions was minimal. The radar’s high sensitivity and large dynamic range point to loss tangent variations as the dominant mechanism for these englacial reflective horizons. The radar is capable of mapping interfaces with reflection coefficients as low as −80 dB near the firn–ice transition and as low as −64 dB at depths of 600 m. We found that firn horizon reflectivity strongly mirrored density variance, a result of the near-unity interfacial transmission coefficients. Zones with differing compaction mechanisms were also apparent in the data. We were able to sound many ice shelves and areas of shallow ice. We estimated ice attenuation rates for a few locations, and our attenuation estimates for the Ross Ice Shelf, West Antarctica, appear to agree well with earlier reported results.

Accelerated loss of ice from Greenland and Antarctica has been observed in recent decades. Ice thickness is a key factor in making predictions about the future of massive ice reservoirs and can be estimated by calculating the exact location of the ice surface and bottom in radar imagery. Identifying the locations of ice boundaries is typically performed manually which is a very time consuming procedure. Here we propose a novel approach which automatically detects the complex topology of ice surface and bottom boundaries based on charged particle concept. Here we first applied anisotropic diffusion to remove the noise and enhance the image. At the second step, we detected the contours in the image based on Coulomb's electrostatic law and the assumption that each pixel is an electrically charged particle. The final ice surface and bottom are detected based on the projection profile of the contours. The results are evaluated on a large dataset of airborne radar imagery collected during IceBridge mission over Antarctica and show promising results with respect to hand-labeled ground truth.

The basal topography of most of the glaciers that drain the ice caps of the Canadian Arctic Archipelago is largely unknown. To measure the basal topography, NASA Operation IceBridge flew a radar depth sounder in a wide swath mode with three transmit beams to image the glacier beds during three flights over the archipelago in 2014. We describe the measurement setup of the radar system, the algorithms used to process the data to produce a 3D image of the glacier bed, show digital elevation model (DEM) results of the beds, and provide a basic assessment of the tracking algorithm used to extract the DEM.

We have developed an unmanned aerial system consisting of a compact sounding radar operating in the frequency bands of 14 and 35 MHz integrated into a fixed-wing UAV for remote surveys of glaciers and ice-sheets. The system is capable of collecting coherent sounding measurements along multiple parallel tracks. With the use of differential GPS for precise trajectory determination, we demonstrate multipass SAR array processing. The system was recently deployed by CReSIS personnel in the spring of 2016 to survey the Russell glacier in Greenland. This paper reports on the instrumentation including the integration of the radar, antennas, and aircraft; the survey flights in Greenland; and results from measurements collected at 35 MHz.

We present an overview of a set of radar instruments developed at the University of Kansas for multi-spectral measurements of ice and snow properties. The systems operate at different frequency bands ranging from 14 MHz to 38 GHz, onboard manned and unmanned aircraft. The data collected with these systems are used to estimate parameters such as ice thickness, ice surface and bedrock topography, snow cover thickness on sea ice, and annual snow accumulation. We give a summary of recent field programs (including operations out of Punta Arenas, Chile) and discuss current collaborations with Chilean institutions.

Radar research has been synonymous with the University of Kansas (KU) for over half a century. As part of this special session organized to highlight significant radar programs in academia, this paper surveys recent and ongoing work at KU. This work encompasses a wide breadth of sensing applications including the remote sensing of ice sheets, autonomous navigation methods for unmanned aerial vehicles (UAVs), novel laser radar capabilities, detection of highenergy cosmic rays using bistatic radar, different forms of waveform diversity such as MIMO radar and pulse agility, and various radar-embedded communication methods. The results of these efforts impact our understanding of the changing nature of the environment, address the proliferation of unmanned systems in the US airspace, realize new sensing modalities enabled by the joint consideration of electromagnetics and signal processing, and greater facilitate radar operation in an increasingly congested and contested spectrum.

We present a compilation of radio-echo sounding (RES) measurements of five radar systems (AWI, BAS, CReSIS, INGV and UTIG) around the EPICA Dome C (EDC) drill site, East Antarctica. The aim of our study is to investigate the differences of the various systems in their resolution of internal reflection horizons (IRHs) and bed topography, penetration depth and capacity of imaging the basal layer. We address the questions of the compatibility of existing radar data for common interpretation and the suitability of the individual systems for reconnaissance surveys. We find that the most distinct IRHs and IRH patterns can be identified and transferred between most data sets. Considerable differences between the RES systems exist in range resolution and depiction of the bottom-most region. Considering both aspects, which we judge as crucial factors in the search for old ice, the CReSIS and the UTIG systems are the most suitable ones. In addition to the RES data set comparison we calculate a synthetic radar trace from EDC density and conductivity profiles. We identify 10 common IRHs in the measured RES data and the synthetic trace. We then conduct a sensitivity study for which we remove certain peaks from the input conductivity profile. As a result the respective reflections disappear from the modeled radar trace. In this way, we establish a depth conversion of the measured travel times of the IRHs. Furthermore, we use these sensitivity studies to investigate the cause of observed reflections. The identified IRHs are assigned ages from the EDC's timescale. Due to the isochronous character of these conductivity-caused IRHs, they are a means to extend the Dome C age structure by tracing the IRHs along the RES profiles.

This paper presents a radar sensor package specifically developed for wide-coverage sounding and imaging of polar ice sheets from a variety of aircraft. Our instruments address the need for a reliable remote sensing solution well-suited for extensive surveys at low and high altitudes and capable of making measurements with fine spatial and temporal resolution. The sensor package that we are presenting consists of four primary instruments and ancillary systems with all the associated antennas integrated into the aircraft to maintain aerodynamic performance. The instruments operate simultaneously over different frequency bands within the 160 MHz-18 GHz range. The sensor package has allowed us to sound the most challenging areas of the polar ice sheets, ice sheet margins, and outlet glaciers; to map near-surface internal layers with fine resolution; and to detect the snow-air and snow-ice interfaces of snow cover over sea ice to generate estimates of snow thickness. In this paper, we provide a succinct description of each radar and associated antenna structures and present sample results to document their performance. We also give a brief overview of our field measurement programs and demonstrate the unique capability of the sensor package to perform multifrequency coincidental measurements from a single airborne platform. Finally, we illustrate the relevance of using multispectral radar data as a tool to characterize the entire ice column and to reveal important subglacial features.

Dynamic ice-sheet models are used to assess the contribution of mass loss from the Greenland ice sheet to sea-level rise. Mass transfer from ice sheet to ocean is in a large part through outlet glaciers. Bed topography plays an important role in ice dynamics, since the acceleration from the slow-moving inland ice to an ice stream is in many cases caused by the existence of a subglacial trough or trough system. Problems are that most subglacial troughs are features of a scale not resolved in most ice-sheet models and that radar measurements of subglacial topography do not always reach the bottoms of narrow troughs. The trough-system algorithm introduced here employs mathematical morphology and algebraic topology to correctly represent subscale features in a topographic generalization, so the effects of troughs on ice flow are retained in ice-dynamic models. The algorithm is applied to derive a spatial elevation model of Greenland subglacial topography, integrating recently collected radar measurements (CReSIS data) of the Jakobshavn Isbræ, Helheim, Kangerdlussuaq and Petermann glacier regions. The resultant JakHelKanPet digital elevation model has been applied in dynamic ice-sheet modeling and sea-level-rise assessment.