For the various airborne radar systems operated by AWI, data acquisition and processing rely on different hardware and software packages. These differing workflows converge into a unified framework after the creation of processed radar data products. In this way, standardized data formats are made available from all radar systems for end users, enabling efficient usage, broad accessibility, and consistent archiving. The fundamental workflow or data flow is illustrated in Fig. .

At the acquisition level of each radar system, data undergo preliminary processing before being written to storage. This step may include horizontal and vertical stacking or signal amplification. The raw data at acquisition are stored as primary data in the AWI tape archive within a protected, write-secure area. At the processing level, the raw data are input to various software packages, e.g., the OPR Toolbox formerly CReSIS Toolbox; for UWB and UWBM data, the FMCW Converter for ASIRAS, ACCU, and SNOW data, and the software Paradigm™ from AspenTech Subsurface Science & Engineering for EMR data. The processing outputs, depending on the software, are saved as primary data products in a dedicated section of the AWI tape archive designed for long-term storage, offering high security against data loss.

To ensure uniform data formats and simplified usability, the processed data products in their diverse formats are integrated into an IDL Data Converter. The output consists of standardized NetCDF files for radar data, geodata such as KML and shapefiles, quicklook images of radargrams, and standardized SEGY files and coordinate files for import into commercial seismic software (e.g. Paradigm™) for post-processing and horizon picking. These converted files are also stored on the AWI Hybrid-NAS-Storage, which offers fast connectivity to the internal network and high performance computing resources for analysis and post-processing.

Furthermore, the standardized radar data are made publicly available at two central repositories. Radar data (NetCDF files), radar profile coordinates (KML files), and quicklook radargram images (JPGs) are archived on PANGAEA . Additionally, radar profile coordinates, quicklook radargrams, and other metadata are forwarded to the Marine Data Portal.

To determine the surface reflection, we use a leading-edge retracker for all radar data products. This retracker applies the TCOG (Threshold Centre of Gravity) algorithm within a specific window, capturing the leading edge with a threshold of 0.8. This approach provides a consistent method for determining surface reflection across all radar products by identifying it at the gradient's ascent toward the maximum.

We choose this approach instead of allocating surface reflection at the local maximum because the maximum can be broad, saturated, and potentially characterized by multiple peaks. In a few cases, the algorithm fails: for example, when sidelobes near the surface are too strong, when the tracking window migrates out of the surface reflection due to significant changes in flight altitude, when strong near-surface reflections occur, or when aircraft roll is large. These cases were manually corrected for EMR, UWB and UWBM data.

The picks of the ice base are provided for several radar data products and was traced using different criteria. For all UWB products with a vertical resolution of 5 m or better, the base reflection was predominantly picked at the local maximum of the base reflection. For the EMR products, which has a range resolution of ∼ 5–50 m, the pick was made at the steepest part of the gradient to the local maximum, as the basal reflection return is very broad for this radar product.

Ice base picks are not consistently provided for all radar data sets. Current and future efforts are being made to revise the ice base picks and ice thickness archive from AWI radar measurements. In the future, each dataset on PANGAEA and in the Radar Viewer will include a link to the corresponding curated ice base pick and ice thickness dataset.

The initial historic reason for surveying the polar ice sheets with radar was to map the bed topography beneath the ice. Since 1994, AWI has conducted extensive radar surveys to determine the ice thickness of the Antarctic and Greenland ice sheets. In Antarctica, these surveys focused mainly on Dronning Maud Land . They have been contributing a considerable amount of data to continent-wide ice thickness and bed topography datasets such as Bedmap and BedMachine Antarctica as well as for modelling studies e.g.,see Sect.  for details and serve as a benchmark dataset for automatic ice-base extraction .

In Greenland, the measured ice thickness primarily covers the northeastern part of the ice sheet and the area surrounding the NorthGRIP ice core . The bed topography derived from a survey with the UWB system led to the discovery of a large impact crater beneath Hiawatha Glacier in northwest Greenland . All ice thickness data together form an essential component of Greenland-wide ice thickness and bed topography products .

The majority of ice thickness measurements, both in Greenland and Antarctica, were obtained from surveys with the EMR radar, but was largely replaced by using the UWB radar since 2016 for most glaciological objectives and projects. For aerogeophysical surveys, the simpler and lighter EMR system remains a standard system for combination with gravity and magnetic sensors due to its generally reduced weight and thus greater range. Additionally, for logistical reasons, it can be advantageous to use the EMR when the UWB system is currently in use in the Arctic or not operational. Testing of a lighter gravimeter alongside EMR in 2022/23 paved the way for gravimetry together with UWB for the first time in the Antarctic season of 2024/25. Despite the weight saving, this combination still offers a shorter range than is possible with the EMR system.

Both radar systems are capable of sounding ice more than three kilometres thick, but the improved configuration and processing capabilities of the UWB system provide better range and along-track resolution of the bed topography. For ice thickness determinations using the EMR radar, long-pulse data were used to identify the base reflection in profiles where (i) short-pulse data were not available or (ii) the base reflection was not visible in the short-pulse data. Since short-pulse data offer better vertical resolution, they were preferred for ice thickness determination where the base reflection is visible due to the higher vertical resolution of the data product.

In addition to studies focusing on ice-thickness determination, AWI radar data have also been used to examine the detailed properties of the ice sheet base. In both Greenland and Antarctica, bed topography analyses have been employed to investigate valley structures and the roughness of the bed using various metrics . These studies range from a contribution of data in DML to continental-scale studies , to local and regional contexts to explore the relationship between ice flow, its direction, and the direction-dependent basal roughness as well as landscape preservation and origin underneath the AIS and GrIS . To further refine our understanding of bedrock roughness on smaller spatial scales, parameters like the waveform abruptness of the base reflection and characteristic side reflections parallel to flight tracks have been analyzed. For instance at the onset region of the Northeast Greenland Ice Stream (NEGIS) off-nadir reflections indicate the presence of elongated subglacial landforms oriented parallel to ice flow and shaped by ice stream activity , and were categorized in a high-resolution DEM of the bed topography as mega-scale glacial lineations .

Next to mapping bedrock topography and subglacial landscapes shaped by glacial activity, radar data are also important for analyzing the transport and storage of subglacial water. High-resolution ice thickness measurements are essential for modelling how subglacial water moves and identifying potential locations where it might pool, such as subglacial lakes . While radar data alone are often not sufficient for unambiguous lake detection, they serve as a crucial foundation for combination with other measurements or models. For example, in western DML, radar data combined with SAR interferometry have shown how subglacial water moves between topographic depressions .

In addition to radar signals at the ice–base interface, the reflection characteristics of basal ice are also valuable for understanding basal properties and processes. High-resolution ultra-wideband (UWB) data from a 2018 survey of Jutulstraumen Glacier revealed embedded point scatterers, which are likely sediment particles instead of liquid water inclusions . Sediments are likely to become entrained when subglacial water freezes to the ice base, influencing not only the mechanical properties of the basal ice during flow but also providing insights into the basal temperature regime and the presence of subglacial water and upstream basal melting .

One of the fundamental reasons why radar surveys are conducted in Greenland and Antarctica is to identify suitable locations for deep ice core drilling . Moreover, radar measurements near core sites offer a way to extrapolate the information from dated ice cores laterally across a wider area. For this purpose, the European Project for Ice Coring in Antarctica (EPICA) pre-site surveys in western Dronning Maud Land used airborne EMR radar measurements between 1994 and 1999 to search for a region with undisturbed layering, a low flow velocity, and sufficient ice thickness for drilling, ideally at a summit or ice divide . Additionally, the bedrock topography played a crucial role in determining a suitable drilling location .

The ice-thickness data also formed the basis for further ice dynamic modelling leading to the selection of the EPICA DML drilling site near 75° S and 0° . Subsequently, Kohnen Station was established in 2001 at the drill site, providing a logistics base for deep ice core drilling of the EDML ice core . To locate a suitable drill site for the oldest Antarctic ice core, used radar-derived ice thickness data to perform 3-D continental ice-sheet modelling.

Moreover, AWI radar data were crucial in identifying promising sites for 1.5-million-year-old ice in Antarctica. AWI's surveys in the Dome Fuji region (2014/15 and 2016/17) revealed detailed subglacial topography, which refined thermokinematic models to predict frozen basal ice locations . In the Dome C region, AWI radar data, combined with other datasets, identified areas with low basal roughness and minimal subglacial water, which are key for preserving ancient ice . These data were integrated with thermodynamic modeling to constrain geothermal heat flux thresholds, pinpointing Dome Fuji and Dome C as ideal candidates for retrieving old ice cores .

In connection with the North Greenland Ice Core Project (NorthGRIP) for drilling in central Greenland between 1996 and 2001 , an extensive grid was flown around the drilling site to map internal stratigraphic layering, which might threaten the continuity of the ice core chronology , and to connect the new site to the existing GRIP ice core . The data suggested that, at the time, a continuous ice core record reaching back to the Eemian period (129–116 ka BP) in the last interglacial could be expected. Moreover, AWI UWB radar data were used together with NASA's Operation IceBridge data to determine a suitable ice core drill site on Müller Ice Cap on Umingmat Nunaat (Axel Heiberg Island), in the Canadian Arctic (Fig. c) to retrieve an undisturbed Holocene climate record .

Internal reflection horizons (IRHs) are crucial for understanding the age–depth structure of ice sheets and can often be precisely dated using ice cores. Prominent ice-sheet wide IRHs are typically caused by the deposits of global bipolar volcanic eruptions and can often be traced in radargrams across large areas of Antarctica and Greenland . However, a wide range of IRHs can also originate from local or non-bipolar volcanic eruptions. At several ice-core sites in Greenland (, at GRIP, , at NorthGRIP, NEEM and EastGRIP) and Antarctica (, at EDML and , at EDC) electromagnetic forward modelling showed that many of the prominent IRHs in Antarctica and Greenland are linked to conductivity peaks in the ice. Consequently, IRHs are present and intercomparable in various radar data products available to the radioglaciology community . Moreover, IRHs represent consistent age markers, serving as archives of the historical development of the Antarctic and Greenland ice sheets. They are essential for reconstructing the mass balance at both the ice surface and base, as well as for understanding the deformation history of the ice, which is currently a major objective of the SCAR Action Group AntArchitecture .

IRHs have been extensively traced and dated in AWI radar data across Dronning Maud Land and central East Antarctica Fig. c. As part of the fourth International Polar Year AWI conducted the first Dome Connection flights to connect major ice domes in East Antarctica in 2007–2008 Talos Dome–Dome Concordia–Vostok–Dome A;. Moreover, IRHs were traced over coastal regions in eastern DML to investigate the dynamic imprints and histories of ice rises e.g., Halfarryggen and Derwael Ice Rise;.

It has been demonstrated that some of the IRHs traced over large areas in radar data from Dronning Maud Land and Central East Antarctica match to similar IRHs in both East and West Antarctica . This represents a significant contribution to Antarctic-wide modelling approaches that use IRHs to calibrate their models e.g., and to develop a three-dimensional, continent-wide age-depth architecture model as for the Greenland Ice Sheet, as pursued by the AntArchitecture initiative .

Furthermore, AWI data have enabled IRHs to be extensively mapped in central and northeast Greenland Fig. b around the NorthGRIP ice core site and at the onset of the Northeast Greenland Ice Stream. When IRHs are mapped in closely-spaced survey grids, they can be used for constructing 3D surfaces of ice sheet structures providing crucial information on ice deformation.

Only a fraction of visible IRHs have been traced to date. The large archive of radar data over the polar ice sheets thus represents a considerable untapped opportunity for tracing IRHs using semi-automatic tracing methods as well as with the help of machine learning .

Besides exploring areas of the ice sheet that are dynamically stable with undisturbed internal stratigraphy, various AWI radar campaigns have focused on dynamic regions in Greenland and Antarctica, particularly with the aim to better understand ice-stream systems. From 2016 through 2022, surveys using the AWI UWB radar primarily targeted the onset region of NEGIS in the vicinity of the EastGRIP ice core drill site and the region of the 79° NG. The high-resolution radar data collected along and across the ice flow direction at NEGIS, as well as over its shear zones , provided fundamental insights into the formation and temporal evolution of this ice stream . The data help underscore the implausibility of the hypothesis linking the fast-flowing NEGIS to exceptional geothermal heat flux . Furthermore the data contributed towards a better understanding of the evolution of large scale folding with implications for ice stream dynamics in northeast Greenland during the Holocene . Additionally, the AWI UWB operated in 2018 and 2022 near EastGRIP in swath mode, enabling a high-resolution reconstruction of bedrock topography at the NEGIS onset Fig. ;.

investigated the basal conditions of West Ragnhild Glacier in DML to characterize its ice flow regime using satellite remote sensing, AWI airborne radar data and ice sheet modelling. In coastal areas in eastern DML, AWI radar data collected over ice rises were used as ice dynamic proxies by interpreting the isochrone geometry beneath the local ice divides. Here, stratigraphic arches, referred to as Raymond Bumps are a metric for the stability of conditions at the ice-divide, which, in turn, is a proxy for the millenial timescale stability of the surrounding catchment. Examples of surveyed ice rises in Dronning Maud include Halvfarryggen , Derwael Ice Rise , and Hammarryggen ice promontory .

Ice shelves, their interactions with the ocean, and their capacity to buttress tributary ice flows have also been investigated with radar data. This includes, for example, the role of smaller ice rumples which increase ice-shelf damage , the use of highly resolved ice-shelf thickness maps to determine ocean-induced melt rates , partitioning of the ice-shelf composition , and the imaging of subglacial water outlets near the grounding zone .

Airborne radar data from AWI surveys have played a central role in advancing our understanding of ice sheet hydrology, from subglacial lakes beneath the Antarctic ice sheet to englacial channels and surface melt features in Greenland. Since the early identification of bright, flat reflectors linked to subglacial lakes , AWI radar systems have been instrumental in detecting both active and stable water bodies beneath ice sheets . While surface elevation changes help identify actively filling or draining lakes , stable subglacial lakes require radar confirmation using criteria such as high basal reflectivity, relative power contrasts, and echo specularity . However, interpretation using these criteria is complicated by uncertain attenuation rates and the presence of basal ice with entrained water . Additional indicators, such as disrupted internal layering or slope changes in radar horizons, can further support subglacial lake identification .

In Greenland, AWI radar campaigns have contributed key insights into surface meltwater refreezing and firn saturation (Fig. ). Increasing surface melt has led to the widespread formation of ice slabs – dense, refrozen layers within the firn that reduce permeability and enhance runoff . During the 2024 Polarmonitor campaign, AWI used UWB radar in high resolution mode to image stratigraphic detail in these slabs (Fig. c). Tracking their expansion is critical for understanding changes in Greenland's water retention capacity and for benchmarking firn hydrology models. Moreover, firn aquifers act as long-term reservoirs of liquid water in southeastern Greenland's accumulation zone and were mapped using AWI radar data as areas showing strong firn-ice reflectors (Fig. b) and an absence of bed echoes due to signal attenuation due to the presence of water . Transects over regions like Køge Bugt demonstrated that detecting changes in aquifer extent is only possible through airborne radar, as these features are invisible in optical or surface elevation data.

In Antarctica, AWI radar data contributed to assess potential subglacial lake locations underneath the East Antarctic ice sheet . Moreover, radar data in combination with remote sensing products and thermal modelling enabled the detection of (i) cascading subglacial water flow via filling and draining of subglacial lakes as well as (ii) freeze-on of subglacial water with entrained sediment at the onset of the Jutulstraumen Glacier.

AWI radar data also revealed the persistence of englacial channels e.g.near the grounding line (GL) of the Roi Baudouin Ice Shelf, as well as channels formed during supraglacial lake drainage at Greenland's 79° NG . Here, a 21 km² lake has repeatedly filled and drained over the past decade and repeated UWB surveys since 2016 revealed englacial features that remain detectable and mobile years after formation . These observations offer rare insight into how long such drainage pathways remain open and how they interact with glacier motion. Additional radar observations have shown that lake ice in supraglacial lakes causes a pronounced thickness gradient, forming continuously at the upstream end as the glacier flows downstream . This internal structure influences lake stability and drainage behavior, with implications for ice flow.

In addition, fractures and crevasses are important for both hydrology and ice mechanics and are well resolved in AWI radar data. Crack tips appear as hyperbolas in radargrams and can be analyzed to assess crevasse depth and geometry. Radar surveys have detected basal crevasses on the Ross Ice Shelf and surface fractures in Greenland . These measurements have been used to estimate energy release and to identify fatigue cracks . Moreover, fractures linked to supraglacial lake drainage are clearly visible in radargrams, offering a new window into meltwater-induced fracturing processes .

AWI airborne radar data play an important role in a number of studies of large-scale geological variability and tectonic structure in Antarctica e.g.,. Given that additional geophysical instruments, such as magnetometers and gravimeters, have been used less frequently in Greenland campaigns alongside radar systems (Figs.  and ), this section will primarily focus on scientific findings from Antarctic campaigns. However, magnetic surveys were conducted in Greenland alongside AWI EMR measurements between 1997 and 2004 e.g.,, with data available in compilations such as the Greenland Magnetic Map GREENMAG;.

Once adjusted for the loading effect of the ice sheet, which requires detailed analysis of coincident gravity anomaly data, the regional height of the upper rock surface can be related to variations in the thickness of the crust. In this way, calculated considerable variability about a relatively thick mean crustal thickness in Dronning Maud Land, relating it to the amalgamation and later breakup of the supercontinent Gondwana by tectonic collisions and extension. They noted also that very thick (∼ 51 km) crust of the Wohlthat Massif appears to be subject to ongoing vertical isostatic motions in response to much more recent changes in the thickness of the East Antarctic ice sheet.

Radar-derived bed heights are essential for understanding and interpreting those components of the gravity anomaly signal that vary along with lithological variability, especially for calculating the Bouguer anomaly for certain areas. These calculations remove the gravity effects of density contrasts between rocks, ice, and air, showing the regional crustal thickness variability. Using this kind of approach, identified long linear density contrasts that mark crustal suture zones: the lines of long-vanished oceans that were destroyed during the incorporation of what is now Antarctica into Gondwana, as well the lower density fill of Jurassic-aged extensional basins close to the coast.

and used radar-derived depths to the glacial bed to map networks of V-shaped valleys that they interpreted as evidence for the presence of ancient preserved fluvial landscapes in the deep interior of DML. noted that some of the valleys appear to have been cut by erosion that was focused onto ancient tectonic structures. They considered the geological system's various roles in the history of sediment routing to the floor of the deep Southern Ocean, and in the long-term topographic evolution of Dronning Maud Land, in particular its seaward-facing great escarpment.

Radar data also play an important role in the interpretation of magnetic field anomaly variability . The depth of the magnetic sources can be roughly estimated on its wavelength spectrum or more simply by comparative lineament analyses. Sources that lie deeper than the radar bed depth imply the presence of intervening layers of sedimentary rocks, whose magnetic susceptibilities tend to be small. In this way, reported on the presence of a major, probably Jurassic-aged, sedimentary basin between the South Pole and Pensacola mountains. and elaborated on the structure of the crustal suture zones around Lützow-Holm Bay and further south in DML, some of which were recognized earlier by and . Here, conclude that it comprises fragments both of billion year-old volcanic island arcs and at least one much older continental fragment. The remnants of the ancient volcanic system today cover an area ∼ 600 km wide and nearly 1500 km long, meaning that whilst active it may have been a feature comparable to today's Antarctic Peninsula .

AWI airborne radar data have been crucial in enhancing our understanding of subglacial and bathymetric features beneath Antarctic ice shelves. Across multiple studies, these data have been employed to infer bathymetric models, assess ice shelf stability, and explore the dynamics of ice–ocean interactions.

used AWI EMR radar data from the ANT 1994/95 season to investigate the mass-balance conditions in the southeastern Ronne Ice Shelf (RIS). Moreover, the radar data contributed to a comprehensive data set of ice thickness of the entire Filchner–Ronne Ice Shelf, with specific focus on grounding lines e.g., of Foundation Ice Stream; and marine ice .

utilized radar data in coastal Dronning Maud Land to model subglacial topography beneath the Ekström, Atka, Jelbart, Fimbul, Vigrid, Borchgrevink and Roi Baudouin ice shelves. By integrating ice-shelf thickness derived from radar data with seismic, gravity, and multibeam bathymetric references, the studies revealed key topographic features such as deep troughs and sills that regulate the entry of Warm Deep Water into the cavities beneath ice shelves. These findings underscore the role of bathymetric features in modulating basal melting and shielding even small ice shelves from oceanic heat. Variations in warm water access between different ice shelves highlights the spatial variability of ice–ocean interactions, thereby providing insights into potential impacts of climate-driven changes in water temperature and circulation e.g.,. Moreover, the refinement of the regional bathymetry over the Nivl and Lazarev ice shelves emphasizes the protective role of bathymetric ridges and are important constraints for glaciological and oceanographic models.

These data also helped to determine subglacial water flow at the grounding line, estimating basal melting rates, and assess ice shelf morphology and stability . On the Fimbul and Jelbart Ice Shelf, AWI radar data helped to constrain the dynamic ice shelf evolution based on their surface and basal structure . Moreover, radar-derived ice shelf and ice rise thickness measurements were key to determine the basal mass balance of Ekströmisen and the evolution of its drainage basin over the last 40 000 years .

In Greenland, the disintegration of the largest floating ice tongue at the 79° N Glacier in Northeast Greenland was studied in greater detail. Using AWI airborne radar data, it was possible to determine significant temporal changes in the geometry of the ice shelf as well as the influencing factors and basal melt rates using an oceanographic plume model .

AWI's radar systems were used to analyze surface mass balance (SMB) and snow accumulation rates in different glaciological contexts. Accumulation distribution was mapped on Pine Island Glacier in West Antarctica with the ASIRAS system, demonstrating identification of annual resolution from IRHs on a regional scale . The integration of these radar-derived layers with ground-based neutron probe measurements facilitated precise dating and density profiling, covering over 2300 km, eventually resulting in SMB estimates. The results highlighted SMB distribution trends, such as the pronounced orographic precipitation shadow effects in specific regions, and provided critical inputs for modelling total mass input to the basin without evidence of a long-term trend in accumulation rates .

Similarly, in Greenland, AWI's use of the ASIRAS provided high-resolution measurements of winter snow accumulation rates in the percolation zone . By detecting subsurface reflection horizons corresponding to previous summer melt surfaces, ASIRAS captured spatial variability influenced by topographic undulations and snow redistribution by winds. This method offered accumulation estimates comparable to field measurements and demonstrated the potential for extended regional and temporal monitoring of accumulation rates similar to the usage of ground-based radar systems in Greenland .

Most studies that focused on SMB reconstruction in DML used the radar stratigraphy from shallow ground-penetrating radar surveys in combination with data from firn cores. However, with its capability to resolve near-surface IRHs when operated in wideband mode (Figs. e and b), the AWI UWB radar supports SMB reconstructions over wider areas and in combination with deep-sounding surveys (Fig. d).

At the onset of the Jutulstraumen Glacier in western Dronning Maud Land, the UWB radar was used to investigate basal reflection units that likely result from the refreezing of basal meltwater originating from further upstream. These units contribute to a positive contribution of basal mass balance in this area .

AWI radar data have significantly advanced ice sheet modelling by providing detailed insights into the internal structure and dynamics of ice sheets. A key application has been to identify potential drill sites for retrieving the oldest ice in Antarctica. By combining radar derived isochrone stratigraphy with 1D to 3D thermo-mechanical ice flow models, were able to refine the age-depth scales and basal thermal conditions in the Dome Fuji region. In combination with ground-based radar data, revealed spatial variability in basal melt rates and areas with stagnant ice, helping to constrain models of basal thermal states and accumulation rates e.g.,.

Radar-derived IRHs also play a critical role in calibrating ice-sheet models in both ice sheets . IRHs provide 3D observational constraints that highlight discrepancies in surface mass balance, basal drag parameterizations and flow dynamics. By validating models with observed internal stratigraphy, it can be demonstrated how localized variations in e.g., geothermal heat flux, subglacial topography, and surface mass balance affect long-term ice-sheet evolution . Moreover, models were used to predict the age stratigraphy of locally accumulated ice on the Roi Baudouin ice shelf for a given set of oceanic and atmospheric boundary conditions .

At the onset of NEGIS, utilized isochrones traced from AWI radar data to constrain a two-dimensional ice flow model, revealing that upstream flow introduces climatic biases in ice cores due to advection of snow deposited under different conditions. By modelling backward particle trajectories, they determined source locations and past accumulation rates, highlighting the significant roles of basal melting and sliding in NEGIS dynamics. Based on the englacial folds at NEGIS , modeled the effect of ice anisotropy and density variations on fold amplification into large‐scale folds during flow. Moreover, used radar data and ice-flow modelling to infer the influence of the mechanical anisotropy of NEGIS (further details in Sect. ).

AWI's radar-derived ice thickness data alone were widely used for ice sheet modelling efforts . While many modelling studies require accurate geometries and hence ice thickness data over larger domains, up to the continental-scale, other modelling approaches are benefiting substantially from ice thickness profiles along glacier streamlines. Modelling shows that precise overflight of a streamline is important for the accuracy of the results. Along-flow modelling approaches require bedrock topography data along the precise streamline of the glacier for two reasons: (i) the influence of the roughness and small scale topography on sliding needs to be represented well and/or (ii) the gravitational load has to be realistic. This may not be given in gridded, interpolated datasets from which a transect is extracted see e.g.,.

AWI radar data have been instrumental in constraining ice material behavior and evaluating glacier dynamics. For both an Antarctic ice shelf and a grounded-floating transition zone , radar transects aligned with flowlines were combined with surface elevation or in-situ motion data. This setup allowed simulations to test different constitutive relations under controlled conditions, excluding lateral effects.

In model validation efforts, AWI radar-derived ice geometry was used to assess the adequacy of the Blatter-Pattyn higher-order approximation compared to full-Stokes simulations . Results showed that when model resolution is coarser than a certain fraction of ice thickness, complex physics offers no advantage, underscoring the need for dense geometry data. The study also emphasized that reliable conclusions can only be drawn when ice thickness is accurately resolved, as shown by contrasting simulations of two adjacent outlet glaciers, only one of which had dense radar profile coverage.

At Greenland's 79° NG, AWI radar data enabled high-resolution viscoelastic modelling of tidal forcing, basal lubrication, and elastic deformation . The study revealed how variations in bed topography influence basal sliding and stress transmission kilometres upstream of the grounding line – demonstrating the critical role of elastic processes and the value of high-resolution radar-derived geometry for capturing short-term glacier dynamics.

The orientation of crystal lattice axes in ice grains, referred to as the COF, significantly affects both the propagation of electromagnetic waves through ice and the mechanical properties of ice. This anisotropy affects the mechanical properties of ice but also influences the propagation of radar waves and is, thus, a key focus in the analysis of AWI radar data. In Dronning Maud Land in East Antarctica, radar data combined with the EDML ice core were used to investigate the effect of COF variations on radar wave propagation . Studies showed that changes in COF, observed at different depths, corresponded to consistent radar reflections . These observations also helped explore the origin of the so-called echo-free zone discussing the role of COF .

In Northeast Greenland at the onset of NEGIS, UWB radar measurements in 2018 were pivotal in improving our understanding of how COF around the EGRIP drill site impacts the ice stream's mechanical anisotropy , influencing the directional flow and deformation properties of the ice (Fig. b). Travel time delays of internal reflection horizons from intersecting radar profiles and the radar beat signatures revealed the effect of COF anisotropy on the ice's hardness during deformation . The study further suggests that compared to isotropic ice, certain parts of the ice stream are significantly harder for along-flow extension and compression, while the shear margins could be softened for horizontal-shear deformation by a factor of two. Subsequent analysis of the characteristic wavelength of folds in cloudy bands of the EGRIP ice core in the cm-scale and folds in radargrams in the km-scale confirms the strong mechanical anisotropy of ice due to the lattice preferred orientation at NEGIS .

AWI's radar data products are published in PANGAEA and are directly available for download. The data are organized by campaign or season as consolidated entries, with datasets linked to official AWI campaign and flight numbers (events), sensor types, and general geographic locations. The dataset titles include fixed terms indicating the Arctic or Antarctic season (e.g., ARK 1998 or ANT 2018/19), the radar system used (e.g., radar data for the EMR system, ultra-wideband (microwave) radar data for the UWB(M) systems, accumulation radar data for the ACCU system, and snow radar data for the SNOW system), as well as the survey region. Searchable keywords cover specific radar systems (e.g., AWI EMR, AWI UWB, AWI UWBM, AWI ACCU, AWI SNOW, ASIRAS), geographic locations, and ice core sites if profiles are near an ice core.

The DOIs for the datasets can be found via (1) the Marine Data Portal, (2) through the “Collection of datasets from AWI's radar systems on ice sheets and glaciers” , or (3) by using relevant keywords in PANGAEA's search interface (e.g., https://www.pangaea.de/, last access: 19 April 2026, to search for “AWI EMR” related datasets). Data downloads are facilitated through the “View dataset as HTML” option, which allows users to download individual files or the complete dataset as a compressed ZIP archive or an uncompressed TAR archive (Fig. ). Related data publications, such as those on ice thickness, bed topography, internal reflection horizons, and basal roughness, are also listed. All data sets and also complementary data sets from expeditions with AWI's polar aircraft and are available via PANGAEA's Expedition Portal (https://www.PANGAEA.de/expeditions/, last access: 19 April 2026 – see links below “Aircrafts” for a comprehensive campaign list and links to campaign data).

The radar products are available in NetCDF format. For a description of the variables in the NetCDF files, we refer readers to the description in the respective PANGAEA entries and to Table . In addition to the primary data matrix, supplementary information is included such as the two-way travel time (TWT) to the ice surface and, where available, the ice base reflection along with metadata like sensor height above ground, along-track distance, and more. KML files for radar profile locations and quicklook images are also provided.

For the EMR, ACCU, SNOW, and ASIRAS systems, one data product is provided. For the UWB and UWBM systems, multiple products are available. UWB data include an unfocused (qlook) and a SAR-focused (standard) product. If the UWB system operated in polarimetric mode (e.g., during the ARK 2022 season at the NEGIS onset), data for all four polarizations (VV, VH, HH, HV) are provided. Similarly, the UWBM datasets include products for all four polarizations.

AWI radar data can be freely used, but only for scientific purposes under a CC-BY-NC license. Any other use must be requested and explicitly approved by the radar data administrators. A requirement for publishing with AWI radar data is to cite the corresponding dataset, available on PANGAEA, this paper, and, if available, a relevant original scientific publication. We also recommend acknowledging the AWI grant IDs associated with the specific campaigns used, as noted in the respective PANGAEA entries. All publications associated with the datasets are linked to the individual PANGAEA datasets. For datasets that do not yet have a PANGAEA entry (e.g., those that are relatively new or under embargo), but are visible in the Radar Viewer, the overarching PANGAEA entry for the collection is referenced . For those cases, and for all other issues, the Principal Investigators of the respective campaigns should be contacted directly.

An example of how to reference AWI radar data in a Data Availability statement could be: “AWI's airborne radar data quicklooks and accompanying metadata are available in the Radar Data over Polar Ice Sheets viewer of the Marine Data Portal (https://marine-data.de/viewer/; Franke et al., 2026 [this paper]). Processed radar data products are available on PANGAEA (10.1594/PANGAEA.972094; ). The radar data from the [Campaign Acronym] campaign can be downloaded here: [PANGAEA DOI] ([PANGAEA Citation])”.

An example of how to acknowledge AWI radar data could be: “We acknowledge support from the AWI [Campaign or Project Acronym] radar campaign via the AWI funding grant AWI_PA_XXXXX”. Ultimately, any other format is acceptable as long as the overarching references mentioned above, the specific campaign reference and funding statements are included.