A Comparative Analysis of Vivaldi Antenna Designs for Autonomous Ground-Penetrating Radar Sensing of Antarctic Glaciers

Abstract

Against the background of observed climate change, which increases the risk of glacier-system degradation and the formation of hidden crevasses, the development of lightweight, wideband, and highly directional antenna systems has become a key factor in ensuring the safety of logistics operations and enhancing the spatial resolution and interpretability of ground-penetrating radar monitoring of near-surface snow–ice layers. The effectiveness of such systems is largely determined by the characteristics of the antenna unit, including the operating frequency band, gain, radiation pattern, weight, and resilience under extreme climatic conditions. The aim of this review is to provide a systematic analysis of modern Vivaldi antenna designs and Vivaldi-based antenna arrays, as well as to assess their prospects for application in X-band ground-penetrating radar systems for probing Antarctic snow-ice media. The paper considers the main types of ground-penetrating radar (GPR) antennas, their advantages and limitations, substantiates the priority of detecting hazardous near-surface inhomogeneities, and analyzes the capabilities of the X-band for the high-resolution identification of these inhomogeneities. Particular attention is paid to modern modifications of Vivaldi antennas, including antipodal, balanced, director-loaded, metamaterial-based, and array configurations. The analysis shows that Vivaldi antennas represent a promising basis for lightweight, wideband, and directional GPR systems; however, they require further improvement in terms of gain enhancement, sidelobe and back-lobe suppression, radiation-pattern stabilization, and adaptation to Antarctic operating conditions. Future research should focus on the development of adaptive and phased Vivaldi arrays, the use of metamaterials, Electromagnetic Band-Gap/Frequency-Selective Surfaces (EBG/FSS) structures, and director elements, the creation of lightweight, frost-resistant substrate materials, the advancement of multi-polarization multiple-input multiple-output (MIMO) systems, and the integration of antenna arrays with synthetic aperture radar (SAR) processing adapted to a multilayer snow–ice medium.

1. Introduction

Radio-electronic wave-based methods and instruments have long demonstrated their effectiveness in various fields of science and engineering, ranging from petroleum analyzers to medical diagnostic systems [1,2,3]. Similar approaches have also been widely adopted in cryosphere research, including the monitoring of Antarctic glacier systems, where non-destructive methods are particularly important for obtaining information on the internal structure of the snow–ice cover and its temporal evolution [4,5].

Research in the Arctic and Antarctic remains a highly relevant scientific area due to the crucial role of polar regions in the global climate system, the conservation of natural resources, the development of international scientific cooperation, and the maintenance of technological readiness for operations under extreme environmental conditions [6,7]. At the same time, one of the practical tasks that directly determines the effectiveness of expeditionary activities in such environments is ensuring safe mobility, which largely depends on operational monitoring of snow–ice cover conditions. Crevasses, fractures, and cavities within the glacier body may be hidden beneath the surface snow layer, posing a serious threat to transport vehicles and expedition personnel [8,9]. Under current climatic conditions, this issue is becoming even more significant. Observed changes in the thermal regime of the atmosphere and ocean in the Southern Hemisphere are associated with ice-mass loss and the degradation of individual glaciers and ice shelves [10,11,12], while surface melting and meltwater accumulation can initiate and accelerate crevasse formation [13,14,15]. This increases the likelihood of hidden fractures and voids forming within the snow–ice body, and consequently raises the risks associated with travel routes used by research teams and vehicles [15,16].

Due to the increasing risks outlined above, it is necessary to implement reliable non-destructive diagnostic techniques incorporating artificial intelligence technologies [17,18,19], which can identify potentially hazardous areas before vehicles reach them, thereby enhancing the reliability and resilience of logistics operations.

One of the most effective tools for addressing this issue is GPR technology. It provides a means of probing the upper layers of snow and ice, enabling the detection of voids, crevasses, and other inhomogeneities within the snow–ice cover [20,21]. Of particular interest is the application of GPR systems in the microwave frequency range, as the broad operating bandwidth provides high, up to centimeter-scale, depth resolution when examining near-surface snow layers [9,22]. High-frequency GPR systems of this type make it possible to identify small-scale features of the snow cover, as well as thin layers of freshly fallen or compacted snow. This significantly increases the information content and level of detail of radar data.

The information obtained from such measurements makes it possible to identify, in advance, route sections associated with an increased risk of vehicle breakthrough into hidden voids or crevasses, as well as impact loads acting on equipment components. Studies carried out during Antarctic expeditions indicate that the use of GPR systems substantially improves the safety and reliability of logistics routes in ice-desert environments [8,23]. Thus, GPR represents a key element of an integrated system for ensuring transport safety in polar regions.

Additional opportunities are offered by the use of compact microwave GPR systems mounted on unmanned aerial vehicles, which enables route scanning ahead of a convoy without exposing personnel to risk. Several studies have reported successful experiments involving the installation of GPR on drones for estimating ice thickness [22], as well as for snow-cover mapping and the investigation of buried ice [24,25]. The developed systems, comprising a compact vector network analyzer, a Vivaldi antenna, a global navigation satellite system (GNSS) module, and an onboard computing unit, with a total mass of approximately 1.6 kg, enable high-resolution ice-thickness mapping with a spatial resolution ranging from several centimeters to 1 m and a relative error not exceeding 2% [22].

The choice of the elementary radiator type within an antenna array deserves particular attention. Conventional horn and waveguide radiators provide high radiation efficiency; however, their weight and dimensions substantially limit their applicability in mobile and airborne GPR systems, especially when deployed on unmanned platforms [26,27]. At the same time, loop, bow-tie, and monopole antennas are characterized by insufficient angular selectivity and a broad radiation pattern, which reduces their sensitivity to small-scale subsurface inhomogeneities [28,29]. Vivaldi antennas combine low weight, compactness, ultra-wideband performance, and directional radiation characteristics [30,31,32], making them a promising basis for the development of compact yet highly directional microwave antennas suitable for operation in Antarctic snow–ice conditions.

The body of available evidence confirms the high relevance of developing specialized wideband microwave antennas for GPR systems intended for operation in snow–ice conditions. However, further progress in this field requires a comprehensive review that would systematize existing solutions, identify unresolved challenges and problem areas associated with the use of microwave antennas, and assess their potential suitability for operation under the extreme conditions of Antarctica.

In the context of global climate change and the increasing risks associated with the degradation of glacier systems, the requirements for the reliability and level of detail of subsurface radar sensing have become significantly more stringent [10,11,12]. This issue is particularly critical for high-frequency X-band GPR systems (8–12 GHz), where antennas are subject to a set of conflicting requirements: wideband operation, high gain, narrow beamwidth, stable characteristics over a broad frequency range, low weight, and resilience under extreme climatic conditions [26,30,32]. Existing Vivaldi antenna designs, despite their evident advantages, do not fully satisfy this combination of requirements, which necessitates a systematic analysis of international research and engineering experience and the identification of further improvement pathways.

Several recent review articles have considered related topics, but their scope differs from that of the present paper. Historical and state-of-the-art reviews of Vivaldi antennas mainly systematize radiator topologies and bandwidth-enhancement techniques without linking these solutions to the constraints of snow–ice GPR operation [32]. Reviews of UWB antennas for GPR and through-wall imaging summarize wideband antenna classes and application scenarios, but they do not specifically analyze Vivaldi-based X-band architectures for Antarctic sensing [33,34]. By contrast, reviews of cold-region GPR focus on glaciological measurements and signal interpretation rather than on antenna-unit design and its electrodynamic trade-offs [35]. Therefore, the existing literature remains fragmented with respect to the combined problem of antenna design, X-band GPR requirements, and autonomous operation in Antarctic snow–ice environments.

In this work, the Antarctic logistics context is used not as a separate geophysical case study, but as a set of engineering boundary conditions for antenna selection: low mass, high directivity, stable wideband impedance matching, suppressed side and back radiation, compatibility with mobile or unmanned platforms, and sufficient resolution for near-surface crevasse and cavity detection. The distinctive contribution of the review is an integrated antenna-oriented synthesis that connects: the mechanical hazard posed by near-surface snow–firn bridges; the penetration and resolution capabilities of X-band probing signals; the comparative performance of Vivaldi modifications and arrays; and the design challenges that must be solved to develop specialized Antarctic GPR antenna systems.

The aim of this review is to provide a systematic analysis and comparative assessment of modern Vivaldi antenna designs and Vivaldi-based antenna arrays, as well as to identify promising directions for their adaptation for use in GPR systems for subsurface sensing of Antarctic snow–ice cover.

The main objectives of the review are as follows:

• To briefly systematize the classical types of GPR antennas, including their advantages and limitations;
• To identify the operational limitations of antennas in snow–ice media;
• To substantiate the priority of detecting near-surface crevasses and cavities from the standpoint of the fracture mechanics of snow–firn bridges;
• To justify the selection of the X-band for GPR applications in Antarctica;
• To analyze modern Vivaldi-based antenna solutions;
• To perform a scientometric analysis of publication activity related to the research topic;
• To identify unresolved challenges and promising directions for further research.

The review proposes a comprehensive conceptual framework that links the analysis of Vivaldi antenna designs with the stringent requirements imposed on GPR systems intended for operation under the extreme climatic conditions of Antarctica.

2. Materials and Methods

2.1. Materials

The empirical basis of the review comprised peer-reviewed journal articles and conference proceedings retrieved from Scopus, IEEE Xplore, MDPI, and Web of Science, as well as patent documents retrieved from Google Patents. The main technical review covered publications published between 2020 and 2026, whereas the scientometric analysis used a broader dataset covering 2002 to 2026 in order to assess long-term publication dynamics and changes in research interest.

To ensure reproducibility, the literature search was structured around three thematic categories. The first category focused on antenna design and included the terms “Vivaldi antenna”, “tapered slot antenna”, “antipodal Vivaldi”, “balanced Vivaldi”, “Vivaldi array”, “gain enhancement”, “sidelobe suppression”, “metamaterial”, “EBG”, and “FSS”. The second category addressed ground-penetrating radar and microwave sensing and included “ground-penetrating radar”, “UWB GPR”, “short-pulse radar”, “X-band radar”, and “synthetic aperture radar”. The third category covered cryosphere and Antarctic applications and included “Antarctica”, “snow–ice medium”, “firn”, “crevasse”, “snow bridge”, and “cold-region GPR”. Boolean combinations of these terms were used to identify records located at the intersection of antenna design, GPR sensing, and Antarctic snow–ice applications.

The inclusion criteria were publication within the main review period, availability of experimental data or a verifiable numerical model, quantitative description of antenna or radar parameters, and direct or transferable relevance to GPR sensing or wideband microwave imaging. Studies were excluded if they lacked quantitative technical parameters, were purely theoretical without antenna or radar system implementation, fell outside the microwave/GPR context, duplicated previously published conference or journal results, or did not provide sufficient methodological detail for comparative assessment.

2.2. Methods

The study was conducted through a systematic review process. A systematic review’s character is defined by using predefined search criteria, explicit inclusion and exclusion parameters, a documented selection process, a fixed set of comparative criteria, and a transparent method for synthesizing results. A meta-analysis could not be conducted because the antenna studies reviewed differ significantly in terms of frequency band, material, aperture size, power delivery system, testing environment, and performance metrics reported; therefore, structured qualitative and quantitative synthesis was chosen as the most suitable approach. The systematic review includes the following procedures [36]:

• Source search and filtering. Targeted collection of literature using predefined keywords, followed by the exclusion of studies that did not meet the relevance criteria.
• Antenna classification. Grouping of antenna types and systematization of their operational limitations in snow–ice media, including weight, radiation pattern, climatic impacts, and contact/non-contact operating modes.
• Comparative and parametric analysis. Quantitative comparison of the characteristics of Vivaldi antennas and their modifications, including antipodal, balanced, director-loaded, metamaterial-based, lens-loaded, hybrid Vivaldi–horn designs, and antenna arrays, using the following metrics: frequency range, gain, impedance-matching level, weight, dimensions, main-lobe width, sidelobe level, and cross-polarization discrimination.
• Tabular summary. Consolidation of the parameters of the analyzed solutions into a summary table.
• Mechanical analysis. Application of a classical stress-distribution model in an elastic half-space to substantiate the priority of detecting near-surface crevasses, including an assessment of the effect of snow–firn bridge thickness on load-bearing capacity.
• Scientometric analysis. Processing of Scopus data for three thematic areas: X-band SAR systems, Vivaldi antennas, and Antarctic geophysical research. Annual publication dynamics, keyword network structure, and geographical distribution were assessed [37].
• Synthesis of conclusions and identification of unresolved challenges. Generalization of the comparative analysis results to formulate target requirements for an antenna array of an X-band GPR system and to identify promising directions for further research.

A brief overview of the research methodology is presented in Figure 1.

3. Design Features and Operational Limitations of GPR Antennas in Snow–Ice Media

3.1. Brief Overview of Classical GPR Antennas

Modern GPR systems generally operate within the microwave frequency range (300 MHz–300 GHz), where the operating frequency determines the trade-off between probing depth and spatial resolution [35]. The classification of antennas by frequency range is presented in Figure 2.

Several main antenna types are predominantly used as transmitting and receiving devices in microwave GPR systems: dipole, loop, and slot antennas, including Vivaldi antennas; bow-tie and spiral antennas; as well as various modifications of horn and planar radiators [33]. The selection of a specific antenna type is determined by a trade-off between the required resolution, probing depth, weight and size characteristics, radiation pattern, antenna time-domain response, and operating conditions [38]. These antenna types are discussed in more detail below.

3.1.1. Horn Antennas

Most standard GPR antennas, especially those operating in the lower part of the frequency range from several hundred MHz to a few GHz, are characterized by considerable weight and large dimensions, which significantly limits their applicability under field conditions [26,27]. For example, wideband horn antennas can provide gains of approximately 20 dBi; however, due to their large aperture dimensions and flared geometry, they are bulky and difficult to deploy [26]. This is particularly critical when operating over loose snow cover, where the weight and dimensions of antenna modules directly affect vehicle mobility and ease of operation.

In recent years, attempts have been made to address this problem through additive manufacturing technologies. For example, studies have demonstrated the feasibility of producing lightweight horn antenna arrays with monolithic waveguide feeding networks using acrylonitrile butadiene styrene-based 3D printing followed by metallization. Such solutions have shown weight reductions of up to 90% compared with commercial counterparts, while providing an array gain increase of up to 3 dB for a $2\times 2$ configuration [39].

3.1.2. Dipole and Loop Antennas

Dipole antennas and their half-wave and elliptical modifications have traditionally been used in GPR systems due to their structural simplicity and compatibility with bistatic configurations, in which the transmitting and receiving elements are separated [40]. The key limitation of dipoles is their size in the lower frequency range. Such antennas can provide greater probing depth; however, the length of a half-wave dipole may reach several meters, making its use on mobile or unmanned platforms impractical [41]. An alternative solution is provided by magnetic loop antennas based on ferrite cores. Studies show that this design can reduce the antenna size by a factor of 25 compared with an equivalent electric dipole, for example, 30 cm instead of 7.5 m at a frequency of 20 MHz, while maintaining performance characteristics and even increasing bandwidth [42]. However, at an operating frequency of 20 MHz, achieving high spatial resolution is fundamentally impossible due to the large wavelength in the medium.

3.1.3. Bow-Tie Antennas

Bow-tie antennas are among the most widely used designs in GPR systems due to their combination of relatively wide bandwidth and ease of fabrication [28]. However, such antennas produce an almost omnidirectional radiation pattern, which leads to the reception of reflections through side and back lobes and, as a result, to a reduction in the signal-to-noise ratio. A comparison of the radiation patterns of a bow-tie antenna and a highly directional antenna is presented in Figure 3. In addition, there is a problem associated with the frequency dependence of the phase-center position. To improve bow-tie performance, resistive loading is used to suppress parasitic ringing after pulse radiation; however, this reduces overall efficiency [43].

3.1.4. Planar and Slot Antennas: Vivaldi and Tapered Slot Antennas

Planar printed antennas, including microstrip patch, slot, and Vivaldi antennas, attract developers due to their compactness and ease of integration. Among them, Vivaldi antennas occupy a special position as exponentially tapered slot antennas. As noted in review studies [32,34], they can provide a compromise between directivity and weight. However, they also have certain limitations. For example, conventional commercially available Vivaldi antennas are characterized by increased sidelobe levels and back radiation, which is critical for GPR applications [34].

Examples of classical printed antenna designs are presented in Figure 4.

3.1.5. PCB Yagi-Uda and Quasi-Yagi Antennas

PCB Yagi-Uda and quasi-Yagi antennas are lightweight, directional antennas that can be manufactured using standard printed circuit board technology [44,45]. They offer structural simplicity, compact planar implementation, and a clear end-fire radiation pattern formed by driven and parasitic elements on the same substrate [45,46]. However, their resonant operating principle limits the simultaneous achievement of a wide continuous 8–12 GHz bandwidth, a stable phase center, low group-delay distortion, and a consistent radiation pattern [44,46,47].

Wideband PCB Yagi-Uda variants usually require additional directors, modified dipoles, baluns, multilayer substrates, or tuning sections [46]. These measures increase the footprint, complicate the feed structure, and increase sensitivity to fabrication tolerances and substrate properties. Therefore, although PCB Yagi-Uda and quasi-Yagi antennas are suitable for compact directional microwave modules, their use in high-resolution Antarctic X-band GPR requires further assessment of impulse response, phase stability, and low-temperature performance.

Thus, the preference for Vivaldi antennas in this study is determined by the combined requirements of Antarctic X-band GPR, rather than by a single antenna parameter. Compared with horn antennas, Vivaldi antennas are lighter and more suitable for compact mobile and unmanned platforms. Compared with bow-tie and monopole antennas, they provide higher directivity and better suppression of off-axis responses. Compared with PCB Yagi-Uda and quasi-Yagi antennas, they offer a more favorable basis for ultra-wideband operation and time-domain waveform preservation. Therefore, Vivaldi antennas represent the most balanced solution for high-resolution near-surface sensing of snow–ice media, where low mass, broad bandwidth, directional radiation, impulse fidelity, and array integration must be achieved simultaneously.

3.2. Operational Limitations of GPR Antennas in Snow–Ice Media

The design advantages of planar printed antennas are accompanied by a number of limitations related to their electrodynamic characteristics. Most planar printed antennas are characterized by relatively low gain values, typically not exceeding 5–10 dBi, which is substantially lower than those of directional horn or slotted-waveguide antennas [34,48,49]. Planar monopole structures often form a nearly omnidirectional radiation pattern [29], which leads to significant energy losses through the side and back lobes of the radiation pattern. As a result, a GPR system records not only reflections from subsurface objects but also a significant number of signals generated by surface topography features, including snowdrifts, crusted snow ridges, irregularities, and areas of variable density. The superposition of these reflections on the useful signal is accompanied by an increase in background clutter and a reduction in the signal-to-noise ratio, as the amplitudes of surface responses become comparable to or greater than those of signals from greater depths. Recent studies show that excessive side and back radiation associated with a broad radiation pattern increases the sensitivity of the system to surface inhomogeneities, thereby limiting the range and reliability of subsurface sensing under snow-cover conditions [50,51].

Additional limitations are associated with the specific features of GPR operation under soft, loose snow-cover conditions. To prevent antenna modules from sinking into the snowpack during contact deployment, transmitting and receiving devices should be positioned at a certain height above the surface [52]. However, most GPR systems are designed for antenna movement directly over ground or ice, for example, on sleds or towed platforms, and are insufficiently adapted to non-contact scanning at heights of several tens of centimeters above the surface [52,53]. In mountainous areas, this creates the need to transport antennas on skis over snow, which introduces additional risks when operating on avalanche-prone slopes.

Thus, even at the stage of analyzing typical design solutions, a discrepancy becomes apparent between most existing GPR systems and the specific operational requirements imposed by snow–ice conditions.

The main discrepancies and limitations of existing antennas include:

• High mass and large dimensions of antenna devices, which substantially limit their use as part of mobile and unmanned platforms [27,53];
• Limited efficiency when operating in loose snow: contact antennas lose stability, whereas the non-contact mode reduces the amplitude of the reflected signal [38,52,53];
• A broad radiation pattern and low angular selectivity, which lead to poorer extraction of target signals against the clutter background [28,38];
• The dependence of vertical resolution on frequency bandwidth: narrowband antennas, such as 500 MHz antennas, provide resolution on the order of tens of centimeters, whereas wideband microwave antennas provide centimeter-scale resolution [36,54];
• Climatic impacts: the electrophysical properties of the snow–ice medium, including moisture content, density, stratification, and temperature, determine the propagation velocity, attenuation, and penetration depth of radio waves, as shown in Figure 5.

For wet snow, propagation parameters are functions of both density and moisture content, the latter being characterized by high spatiotemporal variability, which complicates the interpretation of GPR data [55]. Cyclic changes in the state of the medium, manifested as transitions from frozen to thawed snow, introduce additional variability in radio-wave propagation parameters, causing changes in attenuation coefficients and in the nature of multiple reflections. Experimentally, this has been observed as a factor affecting vertical resolution and probing depth across different seasons [56]. Antenna structures themselves are also directly affected by climatic factors. Temperature variations and thermal cycling lead to changes in the dielectric properties of materials and in the geometry of conductive layers, which can cause frequency shifts, deterioration of impedance matching, and changes in the radiation pattern [57,58]. Additionally, under icing conditions, the local accumulation of snow and ice in the aperture region changes the radiation boundary conditions and can degrade the directional and polarization characteristics. This justifies the use of radio-transparent protective coatings and anti-icing solutions that are designed without noticeable degradation of the radiation pattern [59]. In coastal areas and on sea ice, the situation may be further complicated by the presence of brine-saturated layers, which increase dielectric losses and, consequently, attenuate microwave signals [60].

3.3. Low-Temperature Effects on Dielectric Substrates Used in Printed Antennas

For antennas based on dielectric substrates, the substrate serves not only as a mechanical support but also as part of the electromagnetic structure. At low temperatures and under repeated thermal cycling, changes in relative permittivity, dielectric loss tangent, moisture absorption, thermal expansion, and copper–substrate adhesion may lead to a shift in the impedance-matching bandwidth, variations in group delay, displacement of the effective phase center, and distortion of the radiation pattern. These effects are particularly important for Vivaldi antennas, since the tapered slot operates as a distributed wideband slot structure; even small substrate-induced changes may accumulate along the aperture and modify the impulse response [32,61].

The expected performance characteristics differ significantly depending on the substrate family. FR-4 is an inexpensive and manufacturing-friendly material; however, it exhibits relatively high losses in the microwave frequency range, and its dielectric properties are temperature-dependent. Measurements of FR-4 microwave printed circuit boards during cooling from 300 K to 4 K showed a decrease in the real part of the relative permittivity by approximately 9% and a decrease in the dielectric loss tangent by approximately 70%. However, field operating temperatures in Antarctica are considerably higher than the cryogenic range of 4 K, since even in inland regions, winter temperatures are typically on the order of −50 to −60 °C, while extreme values may fall below −80 °C. Therefore, direct extrapolation of cryogenic data to Antarctic field conditions should be treated with caution [62]. Hydrocarbon–ceramic laminates, such as RO4350B, provide tighter control of relative permittivity and reduce losses in the microwave frequency range. According to the manufacturer, RO4350B has ${\epsilon }_r=3.48\pm 0.05$, $tan\delta =0.0037$ at 10 GHz, a coefficient of thermal expansion along the z-axis of approximately 32 ppm/°C, and a temperature coefficient of relative permittivity of approximately +50 ppm/°C [63]. At the same time, it is important to distinguish between RO4000-series hydrocarbon–ceramic materials and PTFE/ceramic laminates of the RT/duroid family. For example, RT/duroid 6002 has ${\epsilon }_r=2.94\pm 0.04$, $tan\delta =0.0012$ at 10 GHz, and a temperature coefficient of relative permittivity of approximately +12 ppm/°C, whereas RT/duroid 6010.2LM has ${\epsilon }_r=10.2\pm 0.25$, $tan\delta =0.0023$ at 10 GHz, and a pronounced negative temperature coefficient of approximately −425 ppm/°C [64]. These material families may behave differently under temperature variations. A large-magnitude negative temperature coefficient of relative permittivity may cause an undesirable temperature-induced shift in impedance characteristics. Polyimide films are attractive for flexible and lightweight antenna designs because Kapton-type films are intended for operation over an extremely wide temperature range from −269 to +400 °C and retain useful mechanical flexibility. However, flexible antennas require additional mechanical support to prevent deformation of the tapered aperture under wind loading, during transportation, and during operation [65].

From a mechanical standpoint, the main risks during operation under polar conditions include embrittlement, laminate deformation caused by a mismatch between the coefficients of thermal expansion of the copper and dielectric layers, microcracks near metallized edges, and dimensional changes affecting the slot profile. In addition, in coastal regions of Antarctica, where freeze–thaw cycles may occur, moisture ingress can lead to phase transitions that may potentially degrade copper–substrate adhesion and cause local delamination. In inland Antarctica, the air is extremely dry, which makes moisture absorption less critical for sealed or recently deployed antennas [66]. Therefore, the selection of a substrate for Vivaldi antennas should be based not only on relative permittivity and loss tangent values at room temperature, but also on the temperature coefficient of relative permittivity, moisture resistance, coefficients of thermal expansion, copper–substrate adhesion, resistance to cyclic bending, and impact loading. Other factors that are often overlooked include the anisotropy of dielectric properties in materials such as RO4350B and PTFE-based materials, since it may interact with thermal deformation and affect the polarization characteristics of the antenna. Radiative cooling of the snow surface under clear Antarctic skies may enhance low-temperature gradients in the near-surface layer and thereby increase the thermomechanical loading on the materials of the antenna system [67].

For an X-band Vivaldi element or array, hydrocarbon–ceramic laminates or specialized microwave substrates are preferable to conventional FR-4 when stable impedance matching and reproducible radiation characteristics are prioritized. For lightweight deployable arrays, polyimide-based laminates or composite materials may be useful, provided that the frame or radome maintains the aperture geometry. Ceramic substrates provide high dimensional stability and low losses; however, bulk ceramic materials such as alumina or beryllia (beryllium oxide) are brittle and heavy, whereas low-density ceramic composites may be suitable for polar applications, although their cost and processing complexity are higher.

A brief summary of the materials considered is presented in

Table 1. Low-temperature electrical and mechanical behavior of candidate substrate families for Antarctic X-band GPR (created by the authors).

Substrate Family	Electrical Behavior at Low Temperature	Mechanical Behavior	Suitability for Antarctic X-Band GPR
FR-4 glass–epoxy laminate	Low cost, but higher microwave losses and measurable temperature dependence of complex relative permittivity.	Becomes stiffer at low temperature; moisture uptake and thermal mismatch with copper must be controlled.	Acceptable mainly for prototypes or low-cost modules; not optimal for high-stability X-band arrays.
Hydrocarbon–ceramic microwave laminate	Lower microwave loss and tighter relative-permittivity control; suitable for stable 8–12 GHz impedance matching.	Rigid and dimensionally stable; requires careful sealing and mechanical protection.	Promising baseline for rigid Vivaldi elements and phased-array configurations.
PTFE-based microwave laminate	Very low microwave loss and good dielectric stability.	Softness, creep, and fabrication sensitivity may complicate array assembly.	Suitable for high-performance elements when mechanical support is provided.
Polyimide/flexible laminate	Moderate dielectric constant with good thermal endurance; compatible with flexible printed antennas.	High flexibility and low mass; aperture deformation must be prevented by a frame or radome.	Promising for lightweight deployable arrays if mechanical stabilization is ensured.
Ceramic substrate	High dimensional and dielectric stability.	Brittle and relatively heavy.	Useful for laboratory or fixed modules; less attractive for mobile polar systems.

.

3.4. Environmental Adaptability of Pure-Metal and Dielectric-Based Antenna Architectures

For X-band GPR systems operating in the 8–12 GHz range, the environmental adaptability of the antenna architecture must be assessed not only in terms of material robustness, but also with regard to wavelength-scale dimensional tolerances, phase stability, aperture integration, and preservation of the wideband impulse response [68]. Pure-metal antennas, including horns, waveguides, and all-metal tapered-slot structures, have an obvious advantage under polar conditions because their electromagnetic performance is not directly dependent on the temperature-dependent permittivity and loss tangent of the dielectric substrate [31,69,70]. These antennas are generally less sensitive to moisture absorption, freeze–thaw cycling, and substrate embrittlement, while offering high radiation efficiency and high power-handling capability [31,70,71]. Therefore, pure-metal horns and waveguide antennas remain relevant for Antarctic X-band GPR, where maximum mechanical robustness, stable conductivity, and high radiation efficiency are dominant design priorities.

However, X-band operation also imposes system-level constraints that limit the practical use of pure-metal architectures in autonomous or mobile GPR platforms. Although the free-space wavelength is only approximately 25–37.5 mm in the 8–12 GHz band, a horn or waveguide radiator still requires a three-dimensional aperture and a mechanically rigid feed structure [26,69]. This increases the antenna-module height, wind load, snow-accumulation area, and total mass. When a narrow and stable beam is required in both principal planes, the use of horn or waveguide arrays leads to more bulky mechanical assemblies and complicates the integration of feed networks, phase-control systems, shielding, and protective radomes [26,72]. These drawbacks are particularly significant for lightweight GPR systems intended for use on sleds, vehicles, or in autonomous operations over snow-covered surfaces [73].

Dielectric-based printed Vivaldi antennas offer a different set of advantages in the X-band. At 8–12 GHz, the physical dimensions of a printed tapered-slot radiator and the inter-element spacing of an antenna array are compatible with a compact planar PCB implementation [34,74]. This makes it possible to form two-dimensional apertures with a low profile, reduced mass, printed feed networks, integrated directors, reflectors, shielding elements, and multipolarization channels [74,75]. Such features are especially important for X-band GPR, where high range resolution is achieved through a wide bandwidth, while high angular selectivity and reduced clutter require a scalable directional aperture [34,74]. In this sense, the use of Vivaldi antennas is justified not solely by the properties of the individual radiator, but also by their compatibility with lightweight wideband array architectures.

At the same time, the dielectric substrate becomes a critical factor at X-band frequencies. Temperature-induced variations in relative permittivity and loss tangent can shift the impedance-matching band, alter the phase velocity in the feed structure, and distort the radiation pattern [62,63,64]. Since the electrical dimensions of the radiator and feed network are comparable to the wavelength, even small changes in substrate properties or laminate geometry may affect phase coherence across the array [61,74]. Mechanical effects such as embrittlement, microcracking, delamination, and deformation under thermal cycling can additionally degrade the repeatability of antenna characteristics. Therefore, the use of printed Vivaldi antennas in Antarctic X-band GPR requires careful selection of low-loss frost-resistant substrates, edge sealing of the laminate, protection against moisture ingress, and verification under low-temperature thermal cycling [63,64,65,76].

Thus, a practical antenna unit for Antarctic X-band sensing may combine both approaches: a low-loss frost-resistant printed Vivaldi radiator, a metallic back reflector or shielding frame, sealed laminate edges, a radiotransparent radome, and pre-deployment testing under thermal cycling. Such a hybrid architecture preserves the low mass, planar scalability, and wideband array compatibility of printed Vivaldi antennas, while reducing their vulnerability to temperature variations, icing, and mechanical loading in the polar environment [74,75,77].

The combination of the factors listed above indicates the need to develop new antenna designs for GPR systems. Such designs should combine low weight, a wide operating frequency band, high radiation directivity, and stability of electrodynamic characteristics under climatic impacts, while also being adapted to the specific operating conditions of polar regions.

The comparison of various antenna types reveals that none of the traditional solutions can be chosen based on a single performance metric. Horn antennas offer the highest gain and angular selectivity; however, their aperture size and weight are not well suited for lightweight mobile or unmanned platforms. Dipole and loop antennas have a simple structure and can be efficient at lower frequencies, yet their wavelength-limited dimensions and resolution may not be optimal for detecting near-surface defects at the centimeter scale. Bow-tie antennas offer wideband operation and easy fabrication, but their broad radiation pattern increases susceptibility to surface clutter and off-axis responses. Planar slot and Vivaldi antennas best meet the requirements for low mass, wideband operation, and array integration, although they require additional design considerations to enhance gain, stabilize radiation patterns, and reduce sidelobes and back lobes. Therefore, the antenna design problem for Antarctic X-band GPR should be approached as a multi-objective optimization problem rather than as a simple selection of a commercially available antenna type.

The comparison therefore reinforces the rationale for choosing Vivaldi antennas. They are not selected solely on the basis of their lightweight or broadband characteristics, but rather because they offer one of the most favorable combinations of bandwidth, directional radiation, planar integration, array scalability, and acceptable environmental adaptability, provided that an appropriate low-temperature substrate and protective housing are employed.

4. Rationale for Prioritizing Near-Surface Targets and the Capabilities of X-Band Probing Signals for Their Detection

For the appropriate selection of GPR antennas intended for operation in Antarctica, it is first necessary to define priority sounding targets and justify the choice of the operating frequency range. The analysis presented in Section 3 revealed the limitations of existing antennas. This section considers two key aspects: near-surface crevasses and cavities as the primary source of hazard during vehicle movement (Section 4.1), and the penetration capability of the X-band probing signal and its suitability for the sounding objectives (Section 4.2).

4.1. Near-Surface Crevasses and Cavities as the Primary Source of Hazard

To ensure the safe passage of vehicles, sled-tracked tractors, and expedition teams in Antarctica, priority should be given to the detection of near-surface cavities, hidden crevasses, and weakened zones within the snow–firn body [16,23]. This is because the main operational hazard arises not from the mere existence of a void at any depth, but rather from a reduction in the load-bearing capacity of the overlying snow–firn bridge directly within the zone affected by surface loading.

In polar regions, crevasses considered a significant hazard to the movement of personnel and vehicles, and operational Antarctic routes are regularly surveyed using GPR systems in order to identify weak zones and select safe crossing points [16,23].

From a mechanical standpoint, the decisive parameter is the position of the void roof relative to the stress-propagation zone generated by a surface load. In the classical problem of loading an elastic half-space, the stresses induced by a surface load are maximal near the surface and then decrease and redistribute with depth [78]. The region of significant load influence is limited and expands downward. Therefore, when a crevasse or cavity is located close to the surface and a thin snow–firn bridge has formed above it, stresses from a wheel, track, or human step reach the void roof with minimal attenuation, thereby increasing the likelihood of punching failure, shear failure, or local flexural failure of the overlying bridge [79]. Conversely, when the void is located at greater depth, the same load is redistributed over a larger volume of material, while a thick and mechanically cohesive snow–firn cover partially acts as a load-distributing and bridging layer; consequently, the immediate probability of sudden surface collapse is usually lower. The load distribution is presented in Figure 6.

Therefore, for route-safety tasks, it is more important to search not for any deep voids, but for those inhomogeneities that are located within the near-surface load-influence zone and are capable of causing sudden collapse [16,23]. This approach is also consistent with glaciological observations: recent studies show that a significant proportion of crevasses that open and are detected in Antarctic firn are predominantly near-surface in nature, while the strength of porous near-surface firn is significantly lower than that of dense ice [79,80]. In addition, below the near-surface zone, open air-filled crevasses are more susceptible to closure under overburden pressure, and their further opening becomes more difficult as firn density and strength increase [79,80]. Accordingly, in the context of movement safety, shallow and intermediate-depth features are of primary interest, as they directly control the load-bearing capacity of the upper layers of the snow–firn cover.

4.2. Penetration Capability of X-Band Probing Signals and Their Suitability for Subsurface Sensing Tasks

The selection of the operating frequency band for the GPR system is determined by a trade-off between spatial resolution and probing depth [35,54]. To identify near-surface crevasses and evaluate the thickness of snow bridges, the system must provide sufficient penetration depth to detect the void roof, while at the same time ensuring high resolution for locating objects in the load-influence zone [16,23,79].

Fundamental estimates of penetration depth. Recent studies based on X-band radar data acquired in Antarctica show that the penetration depth of a radar signal in cold and dry polar snow–firn cover can reach several meters [81]. At the same time, spatial variations in penetration depth are observed, which are associated with latitude, surface elevation, temperature, and the properties of the snow–firn body. These factors must be taken into account when processing and interpreting radar data [81,82].

Experimental data on crevasse detection. Subsequent studies on the Totten Glacier in East Antarctica have provided direct experimental evidence of the effectiveness of X-band radar systems for detecting hidden crevasses. A comparative analysis of TerraSAR-X satellite imagery (9.6 GHz, X-band) and helicopter-borne GPR data showed that 100% of crevasses beneath snow bridges up to 4 m thick and 95% of crevasses beneath snow bridges up to 10 m thick can be detected in the X-band [83]. These results are directly relevant to route-safety tasks, as typical snow bridges over crevasses that pose a hazard to vehicles have thicknesses within this range [16,23].

The effect of firn structure on the penetration capability of the probing signal. The penetration capability of an X-band probing signal is determined not only by frequency but also by the electrophysical properties of the medium. Studies of the dielectric properties of Antarctic firn show that the real part of the dielectric permittivity, $\epsilon \prime$, increases with depth and density, while the imaginary part exhibits a complex frequency dependence due to the presence of impurities and salts [84]. The most significant limitation to probing depth is the presence of liquid water in the snowpack, which is critical for coastal areas and seasonal melting zones, but not for inland routes in East Antarctica [55]. Dry firn has low dielectric losses, which enable the X-band probing signal to penetrate to meter-scale depths and, under certain conditions, detect crevasses beneath snow bridges up to 10 m thick [81,83,84].

In addition, the bandwidth of the probing signal is of key importance, as it determines the range resolution of the GPR system and, consequently, its ability to distinguish thin layers, closely spaced interfaces, and local inhomogeneities. This is explained by the physical relationship between bandwidth and resolution [33,54]. In GPR systems, the vertical resolution $∆r$ is inversely proportional to the effective frequency bandwidth $∆f$ according to Equation (1):

$$∆r={\displaystyle \frac{\vartheta }{2∆f}},$$

(1)

where $\vartheta$ is the propagation velocity of the electromagnetic wave in the medium.

For a narrowband signal, $∆f$ is small, which results in low resolution, the merging of reflections from closely spaced interfaces, and the inability to reliably distinguish thin weakened zones and local inhomogeneities. These limitations can be overcome by using wideband signals. A comparison of the use of narrowband and wideband probing signals is presented in Figure 7.

Thus, the integration of mechanical hazard analysis with theoretical and experimental data on the penetration capability of X-band probing signals makes it possible to formulate a justified approach to antenna selection. The X-band provides sufficient penetration depth, the required resolution, and direct experimental evidence of its effectiveness for identifying crevasses beneath snow bridges of hazardous thickness. Considering current trends in the development of the antenna market and open research, Vivaldi antennas are becoming increasingly widespread, which requires a more thorough analysis.

5. Modern Vivaldi-Based Antenna Solutions

In recent years, printed wideband Vivaldi antennas have been intensively developed, driven by increasing requirements for the accuracy and interference immunity of radar sensing. Antennas of this type are widely used in radar systems for various purposes, including through-wall radar systems, hidden-object detection systems, and GPR systems. The Vivaldi antenna, also known as a tapered slot antenna, is of particular interest for ultra-wideband radar applications due to its combination of a planar structure, low weight, wide operating frequency band, and high radiation directivity. The exponential profile of the slot taper enables operation over a wide frequency range, while the formation of an end-fire radiation pattern along the plane of the printed circuit board provides relatively high gain values [32,34].

These features make it possible to consider Vivaldi antennas as a basic element of modern antenna systems, including those operating in microwave frequency bands. In particular, X-band radar sensing tasks 8–12 GHz, which impose increased requirements on the operating bandwidth and radiation directivity, have been addressed in a number of studies proposing improved Vivaldi antenna designs that partially or fully cover this frequency band [31,32,34]. Several key studies are considered below, including those devoted to antenna solutions for other microwave bands, in order to assess the possibility of using the results achieved in these areas and the prospects for adapting these results to X-band operation.

5.1. Antipodal and Balanced Modifications of Vivaldi Antennas

One actively developing direction in the improvement of Vivaldi antennas is represented by their antipodal and balanced modifications. In the classical configuration, a single-sided Vivaldi slot antenna is excited by a coaxial line or a microstrip feed connected to the slot on one side of the dielectric substrate. Modern studies increasingly focus on antipodal structures, in which the slot is formed between two tapered conductors located on opposite sides of the dielectric substrate [32,34] (see Figure 8). This topology improves matching with the feed source, broadens the operating bandwidth, and increases radiation efficiency at the lower edge of the frequency range.

Additional evidence for the applicability of Vivaldi elements in compact radar systems is provided by the experience gained in the WISDOM project, developed for the ExoMars rover. Within this project, an ultralight polarimetric antenna system based on Vivaldi elements and operating in the 0.5–3 GHz frequency range was developed [85]. Despite the difference in the target medium, this example is of interest for Antarctic GPR systems, as it demonstrates the possibility of creating a low-mass, wideband antenna system designed for operation under extreme conditions and suitable for obtaining polarimetric information on the subsurface structure of the medium.

For example, Cheng et al. [86] presented a compact dual-polarized Vivaldi antenna with a shared aperture formed by four radiating elements grouped into two pairs. Each pair forms a single-polarized linear array, and the two pairs are arranged orthogonally to form two cross-oriented linear polarizations. The design also includes director elements and longitudinal slots used to increase gain and improve directivity, as well as a metallic back reflector that enhances forward radiation and suppresses the back lobe. According to the measurement results, the fabricated prototype exhibits a gain ranging from 5.5 to 14.8 dBi when operating over a wide frequency range of 0.5–3 GHz. At the same time, cross-polarization discrimination (XPD) of more than 20 dB is achieved, indicating good polarization isolation. Experiments on tree-trunk sensing showed that this solution provides radar images of defects with clearer identification of damaged zones and an improved signal-to-noise ratio compared with a horn antenna [86].

Another example of a modern approach to wideband antenna design is presented in the work of Sutradhar et al. [87], where an ultra-wideband Vivaldi antenna and linear arrays based on it were proposed for operation in the S, C, and X bands. The developed designs cover a wide frequency range from 1 to 12 GHz, which makes it possible to consider them as a universal platform for high-frequency radar systems. According to the numerical simulation results, all configurations, ranging from a single element to $1\times 2$, $1\times 3$, and $1\times 4$ antenna arrays, demonstrate stable impedance matching under the condition $\left|{\mathrm{S}}_{11}\right|<-10$ dB, while retaining the characteristic wideband properties of the basic radiator. The results obtained by the authors indicate the potential of such configurations for application in high-frequency radar systems, including tasks that require operation in the X-band.

5.2. High-Gain Vivaldi Antennas

In addition to the topological modifications discussed above, considerable attention is paid to increasing antenna gain and narrowing the beamwidth. These parameters are enhanced through structural additions in the slot-aperture region, including director elements, metamaterial, dielectric, and phase-gradient aperture overlays. The combined use of such solutions makes it possible, on the one hand, to preserve the wideband nature of the antenna and, on the other hand, to increase gain in the main-lobe direction and form a narrower directional beam, thereby improving the radar sensing range and system sensitivity [88].

One example of implementing this approach is the work of Hu et al. [89], who proposed an ultra-wideband Vivaldi antenna based on artificial electromagnetic materials. Each array element is equipped with a set of metallic director plates placed inside the inner region of the exponentially tapered slot, which makes it possible to induce surface currents and enhance the longitudinal component of the electric field, thereby increasing gain and forming a narrower main lobe of the radiation pattern. An additional contribution to radiation concentration is provided by an H-shaped metamaterial lens, which is a two-dimensional array of artificial electromagnetic cells placed in the aperture plane. This lens enables controlled modification of the wavefront phase, thereby enhancing radiation in the axial direction. According to the measurement results, the proposed design operates in the 0.9–4.0 GHz range with a matching level of $\left|{\mathrm{S}}_{11}\right|\le -10$ dB and demonstrates a maximum gain of approximately 15.2 dBi [89]. Although this antenna is intended for relatively low frequencies, namely the L–S bands, the techniques used are conceptually scalable and can be adapted for antennas operating in higher-frequency bands, including the X-band. When transitioning to other frequency ranges, the scaling of director elements and metamaterial cells should be performed with regard to the wavelength in the target band, ensuring in-phase interaction of radiation and a controlled phase distribution across the aperture. In addition, multilayer or gradient metamaterials may be used to smooth phase distortions over a wide frequency band [90].

An alternative structural modification of the Vivaldi antenna is presented in patent CN115173036B [91], which proposes a design with a system of rectangular grooves formed along the edge lines of the tapered slot and a series of annular director elements placed in the aperture region. The rectangular grooves serve to suppress surface currents propagating perpendicular to the main axis of the slot, thereby increasing the effective path length of the radiating current while preserving the physical dimensions of the antenna. According to the patent description, this leads to an increase in gain and a reduction in the sidelobe level of the radiation pattern. Additional performance improvement is provided by a set of annular director elements placed in the exit region of the aperture. These elements induce currents that redistribute the electromagnetic field and direct it predominantly along the radiation axis. The combined use of the grooved-slot structure and annular director elements makes it possible to form a wideband antenna with improved radiation characteristics, including increased gain, a more stable and focused radiation pattern, and maintenance of compact dimensions. These properties make this configuration promising for high-frequency systems that require compact, highly directional antennas [91].

5.3. X-Band Vivaldi Antenna Arrays

Particular attention should be paid to the X-band frequency range in the context of radar applications, as the corresponding centimeter-scale wavelengths enable the formation of a narrower radiation pattern with limited physical aperture dimensions and, consequently, improve spatial selectivity and observational detail [92]. In addition, the X-band is widely used in systems aimed at high-resolution operation, where broad operating bandwidths are required, including for achieving range resolution and forming high-resolution radar images [93].

Achieving high radiation directivity in the X-band frequency range, which covers 8–12 GHz, generally requires the use of antenna arrays in which printed Vivaldi antennas serve as elementary radiators [87,92]. Planar radiators of this class are highly scalable when combined into arrays and, with proper selection of the array spacing and feeding scheme, make it possible to preserve their inherent wideband performance [66]. The literature presents examples of both planar and conformal antenna arrays based on Vivaldi elements [92,94]. This diversity of configurations enables flexible control over the radiation pattern and gain level, allowing antenna characteristics to be adapted to the requirements of a specific radar task [92,94].

A representative example is the work of Lindvall et al. [75], who proposed a tightly coupled dual-polarized array based on antipodal Vivaldi elements. Owing to the optimized mutual arrangement of the radiators and the overlap of the tapers, an impedance bandwidth at the −6 dB level is achieved over the 3–20 GHz range for both polarizations. Analysis of the reflection coefficient of a single element, $S_{11}$, and the total active reflection coefficient of the array (TARC) shows that tight coupling between the elements makes it possible to shift the lower boundary of the operating band compared with an isolated radiator, thereby extending the effective operating range of the array. According to the simulation results, under in-phase excitation, a compact $3\times 3$ configuration produces directional radiation with a maximum achievable gain of up to 15 dBi within the operating frequency range. In addition, the comparative analysis conducted by the authors showed that an increase in the number of elements and the array aperture leads to a proportional increase in gain [75].

For the formation of a narrow radiation pattern in both planes, the use of two-dimensional antenna arrays based on Vivaldi elements is promising. Based on calculation estimates, a $4\times 4$ array configuration with an inter-element spacing of approximately $0.7\lambda$ provides a main-lobe width of about 12–15° in both the azimuth and elevation planes, resulting in a consistent narrowing of the radiation pattern. With such dimensions, the expected gain is approximately 15–18 dBi at 10 GHz, which corresponds to the effective aperture of such an array. To implement feeding in two-dimensional arrays based on Vivaldi antennas, corporate power-distribution networks based on staged cascading of dividers are generally used [95]. In particular, the microstrip Wilkinson power divider is widely used, providing uniform amplitude-phase excitation and high isolation between the outputs [96]. From a design standpoint, the array can be implemented either on a single dielectric substrate or as a modular assembly consisting of several sections [95,97].

Related studies conducted in the field of millimeter-wave 5G/6G communication systems demonstrate both the technical feasibility of apertures of comparable dimensions and the scalability of Vivaldi-based structures [98]. In particular, Ibrahim et al. [99] presented a compact antipodal Vivaldi antenna integrated into a $2\times 2$ MIMO configuration, achieving a realized gain of up to 12.9 dBi through the use of parasitic director elements. Similar developments in sub-terahertz and millimeter-wave communications confirm that gain enhancement is achieved primarily through a linear increase in the number of elements within the array, which is consistent with estimates based on the concept of effective aperture [98]. Taking into account the scalability of such structures, increasing the number of radiating elements in the array at lower frequencies, including the X-band, makes it possible to achieve the required gain values exceeding 20 dBi.

The active use of X-band and millimeter-wave antenna arrays in telecommunication systems, including 5G/6G unmanned platforms, further confirms the technical feasibility of compact high-gain apertures [98,99]. For GPR applications, this experience is of interest not as a direct transfer of the frequency range or communication architecture, but as evidence of the possibility of developing lightweight, scalable, and directional antenna arrays based on printed-circuit technology.

5.4. Innovative Configurations and Materials for Vivaldi Antennas

In addition to classical and array-based designs, innovative configurations using artificial materials and hybrid solutions have been actively developed in recent years. For example, Cheng et al. [100] proposed a compact Vivaldi antenna intended for use in short-pulse GPR systems. To increase radiation directivity, two artificial-material-based elements were integrated into the design: an artificial-material lens (AML) placed in the aperture region and a sidelobe suppressor (SSR) installed along the lateral edges of the radiator. According to experimental data, the developed antenna provides a wide impedance bandwidth at the −10 dB reflection level over the frequency range from 0.7 to 2.1 GHz, as well as a −3 dB gain bandwidth from 1.0 to 2.1 GHz. The introduction of the AML focuses radiation in the upper part of the operating band, 1.4–2.1 GHz, whereas the use of the SSR improves performance in its lower-frequency region, 0.7–1.4 GHz. On average, this results in a gain increase of approximately 1 dB at lower frequencies and approximately 2 dB at higher frequencies compared with the initial antenna configuration without these elements. In addition, substantial suppression of back radiation was demonstrated, with the ratio of radiation levels in the forward and backward directions being approximately 10:1, together with a decrease in the main-lobe width of the radiation pattern, indicating increased directivity and stronger concentration of radiated energy [100].

A separate research direction is associated with hybrid antenna designs that combine the features of printed Vivaldi antennas and horn radiators. One variant of this approach is considered in the work of Andryushchenko et al. [101], where two hybrid Vivaldi–horn antennas were developed for a bistatic GPR system mounted on board an unmanned aerial vehicle. The antennas were fabricated from thin copper sheet 0.5 mm thick and have dimensions of $95\times 225\times 180$ mm, with a mass of approximately 240 g each, providing an operating frequency range of 0.55–2.7 GHz and a maximum available bandwidth of about 2.15 GHz. The mass of one antenna–horn assembly does not exceed 0.25 kg, making this design promising for use as part of suspended GPR modules on unmanned aerial vehicles and lightweight mobile platforms. The obtained results indicate that the use of thin-sheet exponential structures as an alternative to massive metallic horns is an effective means of reducing antenna weight while preserving its wideband characteristics, which is of particular interest for mobile GPR systems [101].

The reviewed Vivaldi-based solutions reveal several common design trends and trade-offs. Antipodal and balanced topologies primarily improve impedance matching, bandwidth, radiation efficiency, and polarization isolation, but they do not, by themselves, provide horn-level gain. Director-loaded, lens-loaded, metamaterial-based, and grooved-slot configurations improve forward radiation and beam concentration, but their effect is often frequency-dependent and sensitive to fabrication tolerances, substrate losses, and boundary condition changes caused by snow or icing. Array configurations are the most direct route to high gain and narrow beamwidth in the X-band. However, they introduce mutual coupling, feed network losses, amplitude-phase imbalance, wideband beam squint, and increased calibration requirements.

For Antarctic X-band GPR applications, the most promising design pathway is not an isolated Vivaldi element but a modular two-dimensional array of antipodal or balanced Vivaldi radiators supplemented with lobe-suppression elements and wideband feeding or true time-delay schemes. This configuration best matches the target combination of low mass, scalable aperture, wide impedance bandwidth, high forward gain and compatibility with SAR-based processing while also leaving room for mechanical adaptation through lightweight substrates, protective radomes and anti-icing coatings.

6. Results and Discussion

6.1. Assessment of Research Interest in the Study Topic

To assess the relevance of the selected research area and determine the position of the present study at the intersection of several thematic domains, a scientometric analysis of publication activity was conducted using the Scopus database. The analysis considered three interrelated areas: research on X-band synthetic aperture radar systems, the development of Vivaldi antennas, and Antarctic geophysical studies. The inclusion of X-band SAR in the analysis is justified by the fact that, in publications related to the X-band, most studies combine the use of SAR with this frequency range, indicating the effectiveness of their joint application. The proposed approach makes it possible not only to assess the overall interest of the scientific community in the topic under consideration but also to demonstrate that the present study lies at the intersection of radar signal processing, ultra-wideband antennas, and applied polar geophysics.

As shown in Figure 9, the keyword co-occurrence network, compiled on the basis of the methodological analysis of literature sources and the scientometric dataset, illustrates the links between the main areas of the present study, including X-band GPR sensing, high-resolution radar processing, Vivaldi antennas, antenna arrays, and Antarctic geophysical research. The network visualization shows that key concepts related to X-band GPR, Vivaldi antenna, antenna array, snow–ice medium, and Antarctic geophysical research form an interconnected interdisciplinary field in which hardware solutions, signal-processing methods, and applied tasks of snow–ice medium diagnostics closely complement each other.

Figure 10 shows the distribution of scientific publications devoted to three interrelated areas: X-band SAR, Vivaldi antenna, and Antarctic geophysical research. The diagrams demonstrate the scientific community’s interest in issues related to the development of high-resolution radar methods, ultra-wideband antennas, and geophysical studies of the Antarctic snow–ice environment.

For the X-band SAR research area, the most pronounced increase in publication activity is observed. While 50 publications were recorded in 2002, the number reached 1111 by 2025. A particularly noticeable increase is observed after 2020, indicating growing interest in synthetic aperture techniques, high-frequency radar methods, digital focusing algorithms, and modern approaches to improving the spatial resolution of radar data. The value for 2026—875 publications—should not be interpreted as a sustained decline in interest, as 2026 represents an incomplete indexing year.

The Vivaldi antenna research area is also characterized by a notable increase in publication activity in recent years. Unlike the broader field of X-band SAR, this topic is associated with a smaller absolute number of publications, but it demonstrates a clear positive trend. Before 2020, publication activity remained relatively low, whereas in 2022–2025 the number of studies increased markedly, from 52 publications in 2022 to 82 publications in 2025. This confirms the growing interest in Vivaldi antennas as promising ultra-wideband radiators applicable to radar systems, GPR systems, microwave imaging, and compact antenna arrays.

The Antarctic geophysical research area is characterized by more stable and mature publication activity. Unlike the two previous areas, it does not exhibit such pronounced exponential growth. During 2002–2026, the number of publications ranged from approximately 119 to 266 publications per year. The maximum values were recorded in 2014, with 266 publications; in 2018, with 256 publications; and in 2025, with 255 publications. This indicates the stable scientific significance of Antarctic geophysical research, driven by the need to study the cryosphere, glacier systems, snow–firn stratification, and subsurface inhomogeneities.

To identify the countries that have made the greatest contribution to the development of research on the geophysical study of Antarctica, a country-level publication distribution map was compiled, as shown in Figure 11. The results indicate that the leading positions in this field are occupied by the United States, the United Kingdom, and Germany. The United States is the leader in terms of publication output, with 1220 publications, accounting for approximately 25.5% of the total number of publications. The United Kingdom, with 538 publications, and Germany, with 479 publications, are also among the most productive countries. Italy, Australia, China, France, Japan, Spain, India, Russia, New Zealand, Canada, the Netherlands, and Norway also make a substantial contribution to the formation of this research field. This distribution indicates the key role of North American, European, and Asia-Pacific research centers in the development of Antarctic geophysics and confirms the high level of international involvement in this research area. The data on the leadership of the United States, as well as on the key role of the United Kingdom, Germany, Australia, China, and Japan, are consistent with a published bibliometric analysis of Antarctic geophysical research.

A joint analysis of the three areas shows that the topic of the present work is interdisciplinary and lies at the intersection of three scientific and technological trends. The observed growth in X-band SAR publications after 2020 can be interpreted as a consequence of several converging technological developments, including the deployment of compact UAV- and mobile-platform GPR systems, progress in digital focusing and back-projection algorithms, and the growing demand for high-resolution remote and subsurface imaging [22,102,103]. In the Vivaldi antenna domain, this increase in publication activity is associated with the transition from single wideband radiators to modified radiators and array-based designs, including antipodal and balanced structures, director-loaded apertures, artificial-material lenses, and MIMO configurations [95,99,100]. Meanwhile, the Antarctic geophysics research domain remains more stable, reflecting the maturity of cryosphere monitoring programs rather than a lack of relevance. The bibliometric intersection of these three domains is still relatively narrow; this confirms that specialized antenna architectures for autonomous X-band GPR sensing of Antarctic snow–ice media are still insufficiently developed and constitute a significant research gap. Thus, the obtained scientometric results confirm the relevance of developing lightweight, wideband, and highly directional antennas based on Vivaldi designs for autonomous X-band GPR systems intended for subsurface sensing of Antarctic snow–ice cover.

6.2. Results of the Comparative Analysis

Based on the analysis of modern scientific publications and patent sources published between 2020 and 2025, information on the parameters of a number of Vivaldi antennas and their modifications, as well as a hybrid Vivaldi–horn design, was summarized. The results are presented in

Table 2. Comparative characteristics of modern Vivaldi antennas and their modifications reported in the most comprehensive publications.

Antenna Type	Source	Description, Material, Weight/ Dimensions	Frequency Range/ Gain	Radiation-Pattern Characteristics	Notes
Dual-Polarized $2\times 2$ Vivaldi Antenna Array	[86]	Printed antenna on an FR-4 substrate (${\epsilon }_r=4.4$; $tan\delta =0.0025$; $h=1$ mm) with copper metallization; four Vivaldi elements with a shared aperture; six longitudinal slots; three directors, and a metallic back reflector. Mass not specified; electrical dimensions: $0.29\times 0.29\times 0.48{\lambda }^3$.	0.5–3.0 GHz at $\left|S_{11}\right|<-10$ dB. Measured gain: 5.6–14.8 dBi.	Directional radiation for horizontal and vertical linear polarizations; narrowing of the main lobe due to the directors and reflector. $XPD\ge 20$ dB; port isolation > 30 dB.	Dual-polarized configuration for wood sensing; compact and portable design; improved radar image quality compared with a horn antenna.
UWB Vivaldi Antenna and $1\times 2$, $1\times 3$, and $1\times 4$ Linear Arrays	[87]	Printed UWB antenna on a dielectric substrate (type not specified); single element: $75\times 45$ mm; $1\times 2$, $1\times 3$, and $1\times 4$ arrays were modeled based on this element. Array mass and dimensions are not specified.	1–12 GHz at $\left|S_{11}\right|<-10$ dB (simulation). Single-element gain > 3 dBi; for the arrays, gain increases with the number of elements.	Radiation pattern directed along the antenna axis; for the linear arrays, a narrower main lobe is formed and directivity increases.	Configuration intended for the S/C/X bands; all characteristics were obtained by numerical simulation in CST; experimental verification was not performed.
Vivaldi Antenna with Directors and a Metamaterial H-Lens	[89]	Printed four-element antenna on a Rogers RT5880 substrate (${\epsilon }_r=2.2$; $h=1.5$ mm); directors are installed in the aperture region; an H-shaped metamaterial lens is placed at the aperture. Mass not specified; dimensions: $0.72\times 0.72\times 0.86{\lambda }^3.$	0.9–4.0 GHz at $\left|S_{11}\right|\le -10$ dB. Measured gain: 6.7–15.2 dBi; gain increase ≈ 1 dB relative to the prototype without the metamaterial.	Radiation pattern directed along the antenna axis; narrowing of the main lobe and reduction in sidelobes due to the directors and metamaterial; high radiation efficiency.	Intended for wideband GPR systems; artificial materials are used to simultaneously increase gain and improve the radiation pattern.
Modified Vivaldi Antenna with a Grooved-Slot Structure and Annular Directors	[91]	Printed antenna on an F4B substrate (${\epsilon }_r=2.6;$ $tan\delta =0.0009$; $h=0.8$ mm); 16 rectangular grooves are formed in the flared slot; 7 annular director elements are placed at the aperture; feeding is provided by a microstrip line with a fan-shaped matching section. Mass not specified; substrate dimensions: $20\times 27$ mm.	Wideband microwave antenna; according to simulation, $\left|S_{11}\right|<-10$ dB within the operating band. Numerical gain values are not normalized, but an increase in gain relative to an unmodified Vivaldi antenna is claimed.	Rectangular grooves suppress transverse surface currents and reduce sidelobes; annular directors redistribute the field along the radiation axis and increase gain.	Compact wideband radiator positioned for 5G communications and other microwave applications; an example of structural modification of a Vivaldi antenna using combined grooved-slot and director loading.
Tightly Coupled Dual-Polarized $3\times 3$ Antipodal Vivaldi Array	[75]	Tightly coupled array of antipodal Vivaldi elements on an RO4350B substrate ($h=0.254$ mm); two orthogonal polarizations are implemented on perpendicular boards; the feed network is placed on a separate board. Mass not specified; single-element dimensions: $120\times 56.6$ mm; array spacing: 24.1 mm; array dimensions: $120\times 120$ mm.	Impedance bandwidth by TARC ($\le -6$ dB): 3–20 GHz; for one polarization, down to 2.74 GHz. $\left|S_{11}\right|\le -10$ dB bandwidth for the central element: 4.6–19.8 GHz. Simulated gain reaches > 10 dBi at frequencies above 10 GHz.	Radiation pattern directed perpendicular to the array axis under in-phase excitation; tight coupling and overlapping tapers extend the lower edge of the band compared with a single element.	Demonstration of a wideband dual-polarized antenna array; scaling to 4 × 4 and larger configurations makes it possible to obtain a gain of approximately 20 dBi in the X-band.
Miniaturized Antipodal Vivaldi Antenna with Directors	[99]	Antipodal Vivaldi antenna on an FR-4 substrate (${\epsilon }_r=4.4$; $h=1.5$ mm) with an array of parasitic circular and square directors above the aperture. Mass not specified; antenna dimensions: $5\times 2.5\times 1.5$ mm3.	58–62 GHz; $\left|S_{11}\right|<-20$ dB. Baseline AVA gain = 9 dBi; with directors, up to 12.9 dBi.	Radiation pattern directed along the antenna axis; the directors provide gain enhancement, main-lobe narrowing, and sidelobe suppression while maintaining compactness.	Extremely miniaturized millimeter-wave antenna for 60-GHz systems; gain enhancement is achieved using parasitic directors without bulky lenses or metamaterials.
Vivaldi Antenna with an AML and an SSR Sidelobe Suppressor	[100]	Printed Vivaldi antenna on an FR-4 substrate (${\epsilon }_r=4.3$; $tan\delta =0.025$; $h=1.52$ mm) with a 200-mm-long exponential aperture; an artificial-material lens (AML) is placed above the aperture; SSR elements are installed along the sides. Mass not specified; board dimensions: $240\times 121$ mm.	0.7–2.1 GHz at $\left|S_{11}\right|<-10$ dB. −3 dB gain bandwidth: 1.0–2.1 GHz; gain increase of approximately 1–2 dB relative to the original antenna.	Radiation pattern directed along the antenna axis; the AML focuses the field in the upper part of the band, 1.4–2.1 GHz; SSR elements improve directivity in the 0.7–1.4 GHz range; the main-lobe width in the H-plane decreases to 39–40° at 2.1 GHz; sidelobe and back-lobe levels are reduced.	Compact wideband design integrating two artificial-material-based elements; provides enhanced directivity while maintaining small dimensions, reduces residual ringing in the time-domain response in pulsed mode, and improves the effectiveness of short-pulse GPR systems.
Hybrid Vivaldi–Horn Antenna	[101]	All-metal hybrid Vivaldi–horn design fabricated from copper sheet with $h=0.5$ mm; the exponential slot smoothly transitions into a horn flare.Mass: 240 g; dimensions: $95\times 225\times 180$ mm.	0.55–2.7 GHz; operating bandwidth based on $\left|S_{11}\right|$ is approximately 2.15 GHz. Gain data are not specified; the antenna is used in a UAV-mounted GPR system.	Wideband directivity; high stability of the radiation pattern in the near field. The horn increases energy in the aperture direction.	Lightweight horn–exponential antenna for onboard GPR systems; demonstrates the possibility of substantially reducing mass compared with classical horns while preserving wideband properties.

. The table includes single ultra-wideband radiators, linear and two-dimensional antenna arrays, designs integrating artificial materials, and millimeter-wave antennas with director elements. For each solution, the main design and operational characteristics are provided, including substrate type, overall dimensions and weight-and-size parameters, operating frequency range, gain, radiation-pattern features, and specific functional properties.

The results presented in Table 1 show that the considered solutions cover a wide range of frequency bands and applications. Most of the studies are characterized by wideband impedance matching, while improved directivity is achieved either through structural modifications of a single radiator or through the transition to array configurations.

At the same time, an analysis of gain values shows that experimentally confirmed values for single printed Vivaldi antennas and their modifications, including lens-based and director-loaded add-ons, are predominantly limited to the range of 10–15 dBi, which is insufficient to achieve the gain levels provided by conventional horn antennas [34,86,89]. This level can be approached mainly by increasing the radiating aperture [75,92,95]. The use of a greater number of antenna elements, together with the associated feed network, phase matching, and mutual coupling effects, inevitably results in increased structural complexity and a corresponding increase in the weight and size of the antenna system [75,95,96]. These limitations should be taken into account when selecting the antenna architecture for X-band GPR systems. When formulating the requirements, it is necessary to maintain a balance between high directivity, gain, and beamwidth, on the one hand, and constraints on weight, manufacturability, and stability of characteristics over the operating frequency band, on the other hand.

The quantitative comparison also allows us to rank the reviewed configurations based on their suitability for Antarctic X-band GPR. Single printed Vivaldi antennas and compact modified radiator antennas are advantageous in terms of weight and integration, although their gain is typically insufficient for long-range or high SNR operation. Lens-loaded, director-loaded, and metamaterial-loaded designs offer a measurable improvement in forward gain and side lobe suppression, although the reported improvements are often specific to a particular frequency subband and may not be uniform across the 8–12 GHz frequency range. Tightly coupled and 2D Vivaldi array antennas are the most promising options for achieving gain values close to or exceeding 20 decibels, as their effective aperture can be increased while maintaining a planar and modular design. However, this advantage comes at the expense of a more complicated feed network, more stringent phase stability requirements, increased sensitivity to mutual coupling, and the need for calibration across the entire bandwidth.

Therefore, the most efficient configuration for future Antarctic X-band GPR systems is a light-weight two-dimensional, antipodal, or balanced Vivaldi antenna array with controlled element spacing, directional or EBG/FSS-based side and back lobe suppression, and a wideband feed network that can maintain phase coherence across the operating frequency range. This architecture provides the optimal balance between gain, beam width, operating bandwidth, weight, and compatibility with SAR focusing in a multi-layer snow-ice medium.

Based on the conducted review and the results of the comparative analysis, several key unresolved challenges in the field of X-band GPR sensing can be formulated:

• Achieving a gain exceeding 10–15 dBi while maintaining a wide operating frequency band;
• Forming a narrow radiation pattern and ensuring a high level of side- and back-lobe suppression;
• Maintaining a balance between energy-related and directional characteristics, on the one hand, and weight and dimensional constraints, on the other;
• Ensuring polarization consistency, which involves forming radiation patterns with similar shapes in both radiation planes and achieving high cross-polarization discrimination.

6.3. Prospects for Further Research

6.3.1. Development of Adaptive and Phased Antenna Arrays Based on Vivaldi Radiators

The manufacturability and scalability of printed Vivaldi radiators make them a promising basis for phased and adaptive antenna arrays with digital amplitude–phase distribution control. This makes it possible to form apertures of the required size and implement electronic control of the radiation pattern, including beam scanning and adaptive reconfiguration for subsurface sensing conditions. Wideband arrays are characterized by a frequency-dependent shift in the direction of the main lobe due to the use of phase control, which leads to performance degradation at the edges of the operating band [104,105]. In this regard, a key task for further research is to ensure a frequency-invariant main-lobe direction using true-time-delay/quasi-true-time-delay (TTD/quasi-TTD) lines and frequency-invariant beamforming methods [104,105].

6.3.2. Application of Metamaterials, Electromagnetic Band-Gap (EBG) Structures, Frequency-Selective Surfaces (FSS), and Director Elements to Enhance Radiation Focusing and Suppress Side and Back Lobes

The combination of structures of this type has already demonstrated effectiveness in a number of wideband antenna configurations [106,107]. However, ensuring their stable operation over the entire 8–12 GHz range requires further experimental studies and refinement of design methodologies, since the effectiveness of EBG and FSS elements is often limited to narrow spectral regions, which can lead to performance degradation at the edges of the operating band [108].

6.3.3. Development of Lightweight Substrate and Structural Materials

A significant potential for reducing antenna mass while maintaining the required electrodynamic and mechanical characteristics is associated with the use of lightweight substrates and supporting structures [109,110]. This group includes composite and porous dielectric substrates with reduced effective dielectric permittivity and density, multilayer structures with air gaps, as well as flexible polymer materials, such as polyimide [109,110], used as low-mass substrates in combination with lightweight frame-type supporting structures. The application of such solutions makes it possible to reduce the weight of the antenna system without substantial degradation of mechanical strength. This is particularly relevant for airborne and portable GPR systems operating under extreme climatic conditions.

6.3.4. Development of Multi-Polarization and Multichannel MIMO Systems Based on Vivaldi Antennas for Simultaneous Operation in Several Orthogonal Polarizations

For subsurface sensing tasks, including the detection of crevasses, inhomogeneities, voids, and thin interlayers, polarization flexibility becomes particularly important. This makes it possible to increase sensitivity to different types of subsurface structures and to the nature of their anisotropy [111]. Multichannel wireless configurations, in particular MIMO systems [102,112], enable spatio-temporal aperture synthesis, which leads to improved resolution and reduced speckle noise in radar image formation.

6.3.5. Integration of Vivaldi Antenna Arrays with Synthetic Aperture Methods

The technological flexibility and scalability of antennas of this type make it possible to form apertures of the required configuration [113], while SAR processing [103] provides coherent summation of signals recorded during sequential movement of the antenna or in multichannel schemes. This contributes to improved cross-range resolution, sensitivity, and signal-to-noise ratio without a substantial increase in the physical dimensions of the antenna structure. For X-band GPR systems aimed at detecting hidden crevasses, subsurface cavities, weakened firn zones, and other near-surface inhomogeneities in the snow–ice medium, this combination of hardware and algorithmic solutions is particularly relevant. At the same time, classical SAR algorithms based on the assumption of an effectively homogeneous medium have limited applicability in a stratified snow–firn–ice sequence. In this regard, further research should focus on matching the electrodynamic parameters of Vivaldi arrays with the requirements of wideband SAR processing, as well as on developing adapted focusing methods that account for layer-by-layer signal propagation, refraction at interlayer boundaries, phase accumulation, and signal attenuation in a multilayer snow–ice medium [103,114].

7. Conclusions

The conducted review confirmed the high relevance of developing compact, wideband, and highly directional antenna systems for GPR sensing of Antarctic snow–ice media. It was shown that, under conditions of increasing climate-related risks, growing uncertainty in the state of the snow–firn cover, and the need to ensure the safe movement of expedition vehicles, rapid detection of near-surface crevasses, cavities, and weakened zones becomes critically important. These inhomogeneities pose the greatest hazard from the standpoint of the mechanical stability of snow–firn bridges. It was substantiated that the X-band, owing to its combination of sufficient penetration capability in dry polar firn and high spatial resolution, is a promising frequency range for high-detail near-surface sensing tasks.

The comparative analysis of classical GPR antennas and modern Vivaldi-based designs showed that conventional horn, loop, dipole, and bow-tie antennas have a number of limitations related to weight, dimensions, broad radiation patterns, insufficient angular selectivity, and sensitivity to side and back reflections. Against this background, Vivaldi antennas stand out as the most promising hardware basis for mobile and unmanned GPR systems owing to their planar structure, low weight, ultra-wideband performance, and directional radiation characteristics. At the same time, the analysis showed that existing single-element and modified Vivaldi antennas do not yet provide the full combination of required characteristics, including high gain, a narrow and uniform radiation pattern, strong suppression of side and back lobes, low weight, climatic resilience, and parameter stability over the entire 8–12 GHz range.

Thus, further research should focus on addressing a number of open challenges related to increasing gain without a substantial increase in mass and dimensions, forming a stable narrow radiation pattern over a wide frequency band, suppressing side and back radiation, applying metamaterials, EBG/FSS structures, and director elements, developing lightweight frost-resistant materials, and creating multichannel and phased Vivaldi arrays. Of particular importance is the integration of such antenna systems with SAR processing methods adapted to a multilayer snow–ice medium. Solving these challenges will make it possible to move from individual antenna solutions to the development of next-generation specialized X-band GPR systems capable of ensuring high-precision, safe, and rapid detection of hazardous near-surface inhomogeneities under Antarctic conditions.

In summary, the comparative analysis identifies a lightweight, modular, two-dimensional array of antipodal or balanced Vivaldi radiators as the most promising configuration for Antarctic X-band ground penetrating radar systems. The preferred design should combine an aperture that can be scaled to achieve gain values of approximately 20 dB or higher, a wide impedance bandwidth spanning 8–12 GHz, stable main-lobe directionality, side-and back-lobe suppression using director, EBG/FSS or artificial material elements, and compatibility with coherent synthetic aperture radar processing. The main unresolved challenges are achieving high gain and low mass simultaneously, suppressing mutual coupling in compact arrays, stabilizing the radiation pattern across a wide frequency range, preserving time-domain pulse integrity, developing frost-resistant, lightweight substrates and radomes, and calibrating the antenna array in changing snow–ice conditions. Addressing these challenges will provide a technical basis for next-generation autonomous GPR systems capable of high-resolution and reliable detection of hazardous near-surface features in Antarctic snow-ice media.