Applications of Ground-Penetrating Radar in Asteroid and Comet Exploration

Abstract

Nowadays, asteroid and comet exploration is one of the most important components of deep space exploration. Through asteroid and comet exploration missions, it is possible to reveal the history of the formation and evolution of the solar system, to understand the origin and evolution of the planets, and to improve scientific models and instruments. As a payload with the advantages of non-destructive, penetrating, and polarizing characteristics, ground-penetrating radar (GPR) has been widely used in lunar and Mars exploration, and will play an important role in planned asteroid and comet exploration missions. In this study, statistics on asteroid and comet exploration missions, scientific results, and space-based ground-penetrating radar (SB-GPR) utilization are presented for the three phases to date. According to the statistics, SB-GPR will play an important role in future Phase 2 and 3 missions. The focus of this study is on analyzing the mission flow, SB-GPR parameters, scientific objectives, and scientific results of the missions that have carried SB-GPR and those that are planned to carry SB-GPR, including the Hera, Rosetta, Castalia, and Tianwen-2 missions. On this basis, the development trends of asteroid and comet exploration missions, as well as the future development trends of SB-GPR design and signal interpretation, are discussed.

1. Introduction

There have been two peak periods in the development of human deep space exploration. The first peak period was during the Cold War space race between the United States and the Union of Soviet Socialist Republics (USSR), during which a total of 166 exploration missions were flown, the landmark missions being the unmanned lunar sample return and the manned lunar landing. The second peak period was triggered by the discovery of possible ice on the Moon by the Clementine spacecraft launched by the National Aeronautics and Space Administration (NASA) in 1994. This period was marked by the entry of the European Space Agency (ESA), the Japan Aerospace Exploration Agency (JAXA), and the China National Space Administration (CNSA) into the ranks of deep space exploration, as well as the realization of Mars landing exploration and the return of samples from small solar system bodies (SSSBs) [1]. Nowadays, the exploration of SSSBs is one of the most important components of deep space exploration, which mainly includes the exploration of asteroids and comets. Asteroids are celestial bodies that orbit the Sun, are smaller in size and mass than planets and dwarf planets, and are less likely to emit gas and dust [2]. Comets are celestial bodies that orbit the Sun and consist of small particles of frozen material, dust, and rock that can release large amounts of gas and dust for extended periods when close to the Sun [3]. Most asteroids and comets preserve information about the early solar system because their interiors are less evolved [2,4,5]. By studying the composition, structure, and orbital characteristics of asteroids and comets, it is possible to reveal the history of the formation and evolution of the solar system and to understand the origin and evolution of planets such as the Earth [6,7]. In addition, the exploration of asteroids and comets is an important way to test new theories and technologies. Scientific models and instruments can be improved through in situ observations and sample returns.

As exploration missions have progressed, a variety of methods and techniques have been developed. Among these techniques, GPR is one of the most reliable tools for studying the surface layer and internal structure of celestial bodies. The advantages of GPR in the study of celestial bodies are threefold: (1) GPR emits electromagnetic waves that can penetrate the interiors of celestial bodies to study their structure and composition, to obtain information about the formation and evolution of celestial bodies, and to assist in the collection of samples; (2) the processes of exploration by means of GPR do not have a destructive effect on the physical structure or composition of celestial bodies; (3) unlike ordinary media, water and ice respond differently to different polarized electromagnetic waves. It is possible to infer the presence of water or ice inside celestial bodies by analyzing the polarization characteristics of GPR echoes.

The earliest application of GPR was the ground-based ground-penetrating radar (GB-GPR). The human exploration of solar system bodies using GB-GPR began in 1946 when a U.S. army ground station transmitted a radar signal to the Moon and received an echo less than 3 s later. Since then, GB-GPR has become one of the most important tools in the exploration of celestial bodies, and is utilized in the Arecibo Radar, the Goldstone Solar System Radar, and the Green Bank Telescope [8]. Over the decades, GPR has developed into a scientific payload for spacecraft on exploration missions, SB-GPR, which has played an important role in the exploration of the Moon and Mars. In 1972, NASA launched the Apollo 17 spacecraft with two SB-GPRs—Apollo Lunar Sounder Experiment and Surface Electrical Properties—for orbiting and landing explorations of the Moon, respectively. In 2003, the Mars Advanced Radar for Subsurface and Ionosphere Sounding, an SB-GPR launched by the ESA, explored the surface and subsurface structure of Mars and produced a map of the global shallow permittivity distribution of Mars [9]. China has also made breakthroughs in developing SB-GPR technology for deep space exploration missions over the past decade. In 2014, CNSA’s Chang’e 3 spacecraft landed on the Moon and used the Lunar-Penetrating Radar to explore shallow structures on the lunar surface. In 2019, CNSA’s Chang’e 4 spacecraft made the first soft landing on the far side of the moon and also realized the first GPR exploration on the far side of the moon [9]. In 2016, the CNSA launched the Planetary Exploration of China program, which aims to explore the planets of the solar system through a series of missions named “Tianwen”. In 2020, during the first phase of the mission, CNSA launched the Tianwen-1 spacecraft with the Mars Orbiter Subsurface Investigation Radar and the Rover Penetrating Radar to explore the structure of the Mars subsurface by orbiting and landing, respectively [10,11]. In the second phase, the CNSA will launch the Tianwen-2 spacecraft in 2025 to explore the near-Earth asteroid 2016 HO₃ (469219 Kamo’oalewa) and the main-belt comet 311P/PANSTARRS, and the mission will utilize an Asteroid Core Scan Radar (ACSR) to explore the asteroid 2016 HO₃ [12].

In this study, statistics on asteroid and comet exploration missions, scientific results, and SB-GPR utilization are presented for the three phases to date. The focus of the study is on analyzing the mission flow, SB-GPR parameters, scientific objectives, and scientific results of the missions that have carried SB-GPR and those that are planned to carry SB-GPR, including the Hera, Rosetta, Castalia, and Tianwen-2 missions. On this basis, the development trends of asteroid and comet exploration missions, as well as the future development trends of SB-GPR design and signal interpretation, are discussed.

2. Asteroid and Comet Exploration Missions

Asteroid and comet exploration missions have been in development for more than 30 years and have gone through three phases of exploration, ranging from distance to proximity and from surface to interior: (1) Fly-by: Observations of the shape and surface physical properties of asteroids and comets. (2) Surrounding and landing: In-depth understanding and analysis of the surface and interior physical properties of asteroids and comets. (3) Sample return: More in-depth laboratory analysis and study of material samples from asteroids and comets.

The application of SB-GPR is not isolated, but is based on the requirements of a given mission. Early missions were mainly fly-by missions, which, due to long relative target distances, were usually focused on optical payload exploration, such as taking photographs of the target and performing surface spectral analysis. As missions progressed and technology developed, spacecraft were able to get closer to their targets, surround and land on them, and even collect samples of them to return to Earth. In Phases 2 and 3, it is hoped that structural information and physical properties of the internal part of the target can be obtained. The SB-GPR is a payload with the advantage of penetration capability to meet this requirement. The SB-GPR also has non-destructive and polarization characteristics that allow it to also assist in sample collection and analysis during the sampling process.

In this section, we summarize the asteroid and comet exploration missions to date and count them with and without the use of SB-GPR. Based on statistics, a total of four of the five planned missions in Phases 2 and 3 will be scheduled to be equipped with SB-GPR. It is obvious that SB-GPR will play an important role in future Phase 2 and 3 missions. We will give a detailed overview of the missions that have been equipped with or plan to be equipped with SB-GPR in the next section.

2.1. Asteroid Exploration Missions

The total number of asteroid exploration missions to date is 19, including 8 fly-by missions implemented and 3 planned; 2 surrounding and landing missions implemented and 2 planned (1 mission with SB-GPR); and 3 sample return missions implemented and 1 planned (1 mission with SB-GPR). See

Table 1. Asteroid exploration missions.

Phase	Space Agency	Spacecraft	Start Time	Target	With (√)/Without (×) GPR
Fly-by	NASA	Galileo	1991	951 Gaspra	×
			1993	243 Ida	×
		NEAR-Shoemaker	1997	253 Mathilde	×
		Deep Space 1	1999	9969 Braille	×
		Cassini–Huygens	2000	2685 Masursky	×
		Stardust	2002	5535 Annefrank	×
	ESA	Rosetta	2008	2867 Steins	×
			2010	21 Lutetia	×
	CNSA	Chang’e 2	2012	4179 Toutatis	×
	NASA	New Horizons	2019	486958 Arrokoth	×
		Lucy	2023–2033	3548 Eurybates 3548 Queta 15094 Polymele 11351 Leucus 21900 Orus 617 Patroclus and Menoetius 52246 Donaldjohanson	×
	JAXA	Destiny+	2028	3200 Phaethon 2005 UD	×
	NASA	OSIRIS-APEX	2029	99942 Apophis	×
Surrounding and landing	NASA	NEAR-Shoemaker	1998	433 Eros	×
		Dawn	2011	4 Vesta	×
	NASA/ESA	AIDA (DART and Hera)	(2021 and 2024)	65803 Didymos Dimorphos	√
	NASA	Psyche	2029	16 Psyche	×
Sample return	JAXA	Hayabusa	2005	25143 itokawa	×
		Hayabusa 2	2018	162173 Ryugu	×
	NASA	OSIRIS-REx	2018	101955 Bennu	×
	CNSA	Tianwen-2	2025	2016 HO3 (469219 Kamo’oalewa)	√

for details.

2.1.1. Fly-By Missions

The first asteroid fly-by mission took place in 1991, when the Galileo spacecraft launched by NASA on its way to explore Jupiter conducted fly-by explorations of asteroids 951 Gaspra and 243 Ida, and discovered Ida’s natural moons (Figure 1a) [13,14]. Over the next decade, NASA successfully launched four spacecraft, all of which completed asteroid fly-by exploration missions. In 1997, the NEAR-Shoemaker spacecraft flew by the asteroid 253 Mathilde and took 330 images, covering 60 percent of the surface of Mathilde [15]. In 1999, the Deep Space 1 spacecraft returned spectral data from a fly-by of asteroid 9969 Braille, which showed that Braille has a composition roughly equivalent to pyroxene and olivine, with a high geometric albedo, suggesting that its surface is relatively young and non-weathered [16]. The Cassini–Huygens spacecraft flew by asteroid 2685 Masursky at a distance of 1.6 × 10⁶ km on its way to Saturn in 2000 [17]. After completing its exploration of comet 81P/Wild2 in 2002, the Stardust spacecraft made a fly-by of the asteroid 5535 Annefrank, whose main body is shaped like a three-pronged column [18].

Since the turn of the century, ESA and CNSA have each completed their first asteroid fly-by missions. The Rosetta spacecraft launched by the ESA made fly-bys of asteroid 2867 Steins in 2008 and asteroid 21 Lutetia in 2010 on its way to its target comet [19,20]. In 2012, after completing its exploration mission to the Moon, the Chang’e 2 spacecraft launched by the CNSA made a fly-by of and took clear optical images of asteroid 4179 Toutatis [21], as shown in Figure 1b. Recently, NASA has conducted several fly-by missions of specific asteroid systems. The New Horizons spacecraft flew by the binary asteroid system 486958 Arrokoth in the Kuiper belt on its way to explore Pluto and its moons in 2019 [22].

Figure 1. (a) Pictures of 951 Gaspra and 243 Ida taken by Galileo (from NASA/JPL-Caltech); (b) 4179 Toutatis image taken by Chang’e 2 (from [21]); (c) schematic of the NEAR-Shoemaker spacecraft rendezvous with 433 Eros (from NASA); (d) the sample from asteroid Bennu (from NASA/Erika Blumenfeld and Joseph Aebersold).

Figure 1. (a) Pictures of 951 Gaspra and 243 Ida taken by Galileo (from NASA/JPL-Caltech); (b) 4179 Toutatis image taken by Chang’e 2 (from [21]); (c) schematic of the NEAR-Shoemaker spacecraft rendezvous with 433 Eros (from NASA); (d) the sample from asteroid Bennu (from NASA/Erika Blumenfeld and Joseph Aebersold).

Planned missions are as follows: (1) The NASA-led Lucy mission: The scientific objective of the Lucy mission is to conduct fly-bys of seven Trojan asteroids (including a pair of binaries) in Jupiter’s orbit between 2023 and 2033 [23]. (2) The JAXA-led Destiny⁺ mission: The science objective of the Destiny+ mission is to fly by asteroids 3200 Phaethon and 2005 UD in 2028 [24]. (3) The NASA-led OSIRIS-APEX mission: The OSIRIS-APEX mission is an expansion of OSIRIS-Rex. It has the scientific objective of flying by the asteroid 99942 Apophis in 2029, which has a potential impact hazard to Earth [25].

2.1.2. Surrounding and Landing Missions

The two surrounding and landing missions that have been conducted were led by NASA. NEAR-Shoemaker was the first spacecraft to conduct a surrounding and landing mission. NEAR-Shoemaker surrounded and explored asteroid 433 Eros in 1998 (Figure 1c), returning more than 160,000 surface images [26], and successfully landed on the surface of Eros in February 2001, during which it took 70 images [27]. The Dawn spacecraft surrounded and explored asteroid 4 Vesta in 2011, and by analyzing the Howardite–Eucrite–Diogenite meteorites, the researchers concluded that Vesta was a complete protoplanet from the early stages of the formation of the solar system [28].

Planned missions are as follows: (1) NASA and the ESA jointly led the Asteroid Impact and Deflection Assessment (AIDA). AIDA began in 2021 and was divided into two parts. The NASA part is the Double Asteroid Redirection Test (DART), with the scientific objective of autonomously navigating an impact on Dimorphos, the moon of the near-Earth asteroid 65803 Didymos [29]. The ESA part is the Hera mission, with the scientific objective of observing changes in the physical and dynamical properties of the binary asteroid before and after the impact, and Hera will also be the first mission to carry SB-GPR for asteroid exploration [30]. (2) The NASA-led Psyche mission: The scientific objective of the Psyche mission is to perform a long-range surround and explore of the main belt asteroid 16 Psyche in 2029 [31].

2.1.3. Sample Return Missions

The Hayabusa and Hayabusa 2 spacecrafts launched by the JAXA completed consecutive sample returns to asteroid 25143 Itokawa and asteroid 162173 Ryugu in 2005 and 2018, respectively [32,33]. In 2018, the OSIRIS-Rex spacecraft launched by NASA completed a sample return to asteroid 101955 Bennu. Scientists have found evidence of both carbon and water in initial analyses of this material (Figure 1d) [34]. The CNSA plans to launch the Tianwen-2 spacecraft in 2025 to return samples from the asteroid 2016 HO₃ (469219 Kamo’oalewa) [12].

2.2. Comet Exploration Missions

The total number of comet exploration missions to date is 14, including 9 fly-by missions implemented; 2 surrounding and landing missions implemented and 2 planned (3 missions with SB-GPR); and 1 sample return mission implemented. See

Table 2. Comet exploration missions.

Phase	Space Agency	Spacecraft	Start Time	Target	With (√)/Without (×) GPR
Fly-by	NASA	International Cometary Explorer	1985	21P/Giacobini-Zinner	×
	USSR	VeGa 1	1986	1P/Halley	×
		VeGa 2
	JXSA	Suisei
		Sakigake
	ESA	Giotto	1992	26P/Grigg–Skjellerup	×
	NASA	Deep Space 1	2001	19P/Borrelly	×
		Deep Impact	2010	103P/Hartley 2	×
		Stardust	2011	9P/Tempel 1	×
Surrounding and landing	NASA	Deep Impact	2005	9P/Tempel 1	×
	ESA	Rosetta	2016	67P/Churyumov–Gerasimenko	√
	ESA	Castalia	2028	133P/Elst-Pizarro	√
	CNSA	Tianwen-2	2032	311P/PANSTARRS	√
Sample Return	NASA	Stardust	2004	81P/Wild	×

for details.

2.2.1. Fly-By Missions

In 1985, the International Cometary Explorer spacecraft launched by NASA (Figure 2a) flew by the comet’s tail within about 7800 km of the nucleus of comet 21P/Giacobini–Zinner [35]. Subsequently, in 1986, on the occasion of the return of comet 1P/Halley, the “Halleyarmada”—five spacecraft consisting of VeGa 1 and VeGa 2 launched by the USSR, Suisei and Sakigake launched by the JAXA, and Giotto launched by the ESA—flew by and explored Halley [36,37,38]. In 1992, the Giotto spacecraft flew by the second comet, 26P/Grigg–Skjellerup, after surviving the harsh environment around Halley’s Comet and taking a clear image of the nucleus of Halley (Figure 2b) [39].

In the new century, NASA launched three spacecraft in a decade to explore comets. In 2001, the Deep Space 1 spacecraft observed the nucleus of comet 19P/Borrelly during its fly-by with a micro integrated camera and spectrometer [40]. After separating from the impactor, the Deep Impact spacecraft operated in a heliocentric orbit for three years and successfully flew by and explored comet 103P/Hartley 2 in 2010 [41]. In 2011, the Stardust spacecraft flew by comet 9P/Tempel 1, which was the target of Deep Impact. Stardust measured dust and took high-resolution images of the area around the impact site, extending the image coverage to nearly two-thirds of the comet’s nucleus surface [42].

2.2.2. Surrounding and Landing Missions

In 2005, the Deep Impact spacecraft made the first hard landing on Comet 9P/Tempel 1 [43], shown in Figure 2c. In 2016, the Rosetta spacecraft launched by the ESA successfully surrounded and landed on comet 67P/Churyumov–Gerasimenko (Figure 2d) and completed the first exploration of the surface and subsurface structure of the comet nucleus using SB-GPR [44].

Planned missions are as follows: (1) The ESA-led Castalia mission: The scientific objective of the Castalia mission is to surround comet 133P/Elst–Pizarro with two SB-GPRs to characterize the surface and internal structure of the comet [45]. (2) The Tianwen-2 Mission: Surrounding and exploring 311P is also the scientific target of the Tianwen-2 mission [12].

2.2.3. Sample Return Missions

Stardust spacecraft collected cometary dust samples from a coma during a fly-by of comet 81P/Wild, completing the first sample return from a comet [46].

3. Missions Equipped with SB-GPR

There are four asteroid or comet exploration missions that carry or plan to carry SB-GPR: (1) The Hera mission: The Hera mission will carry a low-frequency monostatic synthetic aperture radar (SAR) called JuRa, which will be used to explore the DART impact site and the internal structure of Dimorphos. (2) The Rosetta mission: The Rosetta mission carried an SB-GPR called Comet Nucleus Sounding Experiment by Radio wave Transmission (CONSERT), which performed a global survey of the nucleus of comet 67P, providing important information about the internal structure and composition of the nucleus. (3) The Castalia mission: The Castalia mission is a continuation of Rosetta, which is planned to carry two SB-GPR radars at different frequencies, SOUnding Radar for Comet Exploration (SOURCE) and Shallow Subsurface Radar (SSR), to explore the surface and subsurface structure of comet 133P. (4) The Tianwen-2 mission: The Tianwen-2 mission will carry a two-channel inverse synthetic aperture radar (ISAR) called ACSR, which will be used to explore the internal structure of asteroid 2016 HO₃ and comet 311P, as well as to assist sample return from 2016 HO₃. The parameters of the SB-GPR used in the above mission are shown in

Table 3. Asteroid/comet exploration radar.

Mission	Hera	Rosetta	Castalia		Tianwen-2
Start time	2024	2016	2035		2026 and 2033
Target	65803 Didymos Dimorphos (Asteroid)	67P/Churyumov–Gerasimenko (Comet)	133P/Elst–Pizarro (Comet)		2016 HO3 (469219 Ka-mo’oalewa) (Asteroid) 311P/PANSTARRS (Comet)
Radar	JuRa	CONSERT	SOURCE	SSR	ACSR
Antenna	Dipole	Dipole	Dipole	Vivaldi	Dipole	Vivaldi
Center frequency (MHz)	60	90	20	300–800 (Extended mode: 3 GHz)	150	900
Bandwidth (MHz)	20	10	10	--	40	1200
Maximum exploration depth (m)	100	100	1500	50	50 (Asteroid) 300 (Comet)	5 (Asteroid) 30 (Comet)
Resolution (m)	10	20	7	1	5	0.25
Polarization	Tx: 1 Circular Rx: OC and SC	Tx: 1 Circular Rx: OC and SC	--	Tx: 1 Circular Rx: OC and SC	Tx: 1 Circular Rx: OC and SC	--
Transmit power (W)	10	Orbiter: 4 Lander: 10	10	20	10	10
Signal modulation	BPSK	BPSK	Chirp	Stepped frequency	Stepped frequency	Stepped frequency

.

3.1. Hera Mission

The Hera mission is the European part of the NASA/ESA AIDA mission, which was originally called the Asteroid Impact Mission (AIM) [30]. The AIDA mission flow is shown in Figure 3. The AIDA mission targets the near-Earth binary asteroids 65803 Didymos, with a diameter of 800 m, and Dimorphos, with a diameter of 160 m. The NASA part is the DART mission, which was launched in November 2021 and successfully impacted and deflected Dimorphos in 2022 [47]. Hera will be launched in October 2024 and will rendezvous with the binary in December 2026 to observe the collision site [48]. The scientific objective of Hera is to measure the physical properties of Didymos and Dimorphos, the physical properties of the impact craters, and the efficiency of impact-generated momentum transfer, which will provide important data for the science of asteroids and the evolutionary history of the solar system [30,48].

As shown in Figure 4a, in addition to the orbiter, the Hera spacecraft is equipped with two 6U (10 × 20 × 30 cm³) CubeSats, Juventas and Milani, for auxiliary observations [48,49]. The scientific objectives of Juventas are to determine the gravity field, internal structure and surface properties of Dimorphos. The scientific objectives of Milani are to map the global composition of Didymos and Dimorphos, to characterize their surfaces, to assess the DART impact effect and support gravity field measurements, and to characterize the dust clouds around the Didymos system [48].

Juventas is equipped with a low-frequency monostatic SAR named JuRa [48,49,50]. JuRa inherited and redesigned the radar development framework from the AIM mission [50,51], which was derived from the CONSERT radar from the Rosetta mission [49,50]. As shown in Figure 4b, the JuRa is equipped with four foldable 1.5 m dipoles. It uses a binary phase-shift keying (BPSK)-modulated signal, transmits a circularly polarized signal, and receives two perpendicularly polarized signals. JuRa has a center frequency of 60 MHz, a bandwidth of 20 MHz, a maximum penetration depth of 100 m, and a baseline resolution of 15 m (see Table 3 for detailed parameters) [48,49,52]. Due to the relatively low velocity of the Juventas track, JuRa accumulates multiple signals to improve the signal-to-noise ratio (SNR) [49]. JuRa will be observing Didymos and Dimorphos in April 2027 [48].

The scientific objectives of JuRa are as follows: (1) Exploring the dimensions of the Dimorphos blocks, the size of voids, and the presence of fine dust or gravel filling the voids. Significant spatial variations in the internal geological structure will be identified by spatial variations in the received signals, which will provide constraints for the mechanical modeling of the DART impact process. (2) Estimating the average permittivity of the binary and mapping of its spatial variations in order to obtain information on its composition and porosity, especially in the region of the impact crater, including the realization of a partial penetration of Didymos and the estimation of the permittivity, as well as a complete penetration of Dimorphos and the calculation of the three-dimensional distribution of the permittivity [48,50,51].

3.2. Rosetta Mission

The Rosetta mission, an ESA-led comet surrounding and landing mission, was successfully launched in March 2004, and the mission flow is shown in Figure 5. During Rosetta’s journey to the comet, the spacecraft flew by the asteroids 2867 Steins in 2008 and 21 Lutetia in 2010 [19,20]. In 2014, the Rosetta spacecraft arrived at comet 67P/Churyumov–Gerasimenko and began its exploration.

The Rosetta spacecraft consists of an orbiter (Figure 6a) and a small lander named Philae (Figure 6b). The scientific objectives of the Rosetta mission were as follows: (1) to determine the dynamic, chemical, and physical properties of the nucleus of comet 67P and the interrelationships between the materials of which it is composed; (2) to study the evolution of the comet’s activity; (3) to observe the changes in its orbit under the action of solar wind and to place the lander on the nucleus for in situ analysis [53].

Rosetta was equipped with a low-frequency bistatic radar called CONSERT with a center frequency of 90 MHz and a bandwidth of 10 MHz. The CONSERT antennas are two orthogonal dipoles. They transmit a BPSK-modulated circularly polarized signal and receive two perpendicularly polarized signals. CONSERT can penetrate to a maximum depth of 100 m and has a resolution of 20 m (see Table 3 for detailed parameters) [54]. The scientific objectives of CONSERT were as follows: (1) to measure the average permittivity by the time delay of waves passing through the internal structures of the nucleus (e.g., ice and rock); (2) to constrain the composition of the comet (materials, porosity) by modeling; (3) to determine the large-scale structure of the comet (tens of meters) as well as the irregularities, hierarchical structure, and three-dimensional morphology of the nucleus at smaller scales.

The CONSERT radar penetrated the nucleus by transmitting electromagnetic waves between the orbiter and Philae. The quantities measured were the transit time and the signal amplitude, where the transit time depends on the real part of the permittivity and the imaginary part of the permittivity (related to the conductivity) affects the signal amplitude. Thus, CONSERT data can provide information about the permittivity and spatial structure of the nucleus [55].

The plan was for Philae and the orbiter to be visible to each other after landing so that the signal could be calibrated only by vacuum transmission. Then, as the orbiter moved, the nucleus rotated and the geometry changed, resulting in an occultation (the nucleus was between Philae and the orbiter) that allowed waves to be transmitted through the nucleus.

Unfortunately, Philae’s final landing position, direction, and status were unknown due to the multiple bounces it experienced prior to landing. The link between the orbiter and Philae was broken at 17:31 UTC, meaning that Philae and the orbiter were already in occultation when the measurement began and CONSERT was unable to calibrate with the orbiter in time. The relative positions of the orbiter and Philae were constantly changing due to the orbital motion of the orbiter and the rotation of the nucleus. Finally, with the assistance of CONSERT, the orbiter and Philae were calibrated. CONSERT was operated for about 9 h and received strong signals for about 30 min (18:56 to 19:22 UTC) and 80 min (02:47 to 04:06 UTC) at the beginning and end of the acquisition sequence, respectively. Outside of these two periods, much lower SNR was observed (Figure 7) [55].

According to measurements made by Rosetta, the nucleus of 67P is very homogeneous and has a low average density, which implies a very high porosity (up to 75% to 85%). The nucleus of 67P consists of highly porous material, which consists of ice (mainly H₂O, CO, and CO₂) and dust (mainly silicate material and non-volatile macromolecules), with dust/ice volume ratios ranging from 0.4 to 2.6 [55,56].

The researchers hypothesized that both the real and imaginary parts of the permittivity of the nucleus would be low. Therefore, it is reasonable to assume that the wave propagation path inside the nucleus is almost a straight line. The researchers then simulated signal transmission between the 243 possible landing sites of Philae and the orbiter using a three-dimensional model of the nucleus. The average permittivity of the nucleus was calculated using the following equation [54]:

$${\mathrm{t}}_{\mathrm{R}\mathrm{X}}=({\mathrm{t}}_{\mathrm{T}\mathrm{X}}+\frac{{\mathrm{R}-\mathrm{R}}_1}{\mathrm{c}})\frac{{\mathrm{R}}_1\sqrt{\mathsf{\epsilon }}}{\mathrm{c}}$$

(1)

where t_RX is the receiver time read by the absolute clock, t_TX is the time the wave left the transmitter read by the absolute clock, ε is the average permittivity, R is the total length from the orbiter to the lander, R₁ is the length of the wave traveling through the nucleus, and c is the speed of light. The real part of the average permittivity of the nucleus of 67P was finally determined to be 1.27 ± 0.05, and the extent of Philae’s landing area was narrowed to within 21 × 34 m² [55]. According to the Maxwell–Garnet three-phase mixture model at high porosity, the nucleus consisted of a layered sphere with a dust sphere (ε₂, f₂) covered by ice (ε₁, f₁) and embedded in a vacuum ε_e. The approximate range of permittivity of nuclear constituents can be calculated by the following formula [56]:

$$\mathsf{\epsilon }={\mathsf{\epsilon }}_{\mathrm{e}}+3{\mathsf{\epsilon }}_{\mathrm{e}}\frac{\frac{{\left({\mathrm{f}}_1+{\mathrm{f}}_2\right)}^2\left({\mathsf{\epsilon }}_1-{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2+{2\mathsf{\epsilon }}_1\right)+{\mathrm{f}}_2\left({2\mathsf{\epsilon }}_1+{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2-{\mathsf{\epsilon }}_1\right)}{\left({\mathrm{f}}_1+{\mathrm{f}}_2\right)\left({\mathsf{\epsilon }}_1+2{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2+{2\mathsf{\epsilon }}_1\right)+{2\mathrm{f}}_2\left({2\mathsf{\epsilon }}_1-{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2-{\mathsf{\epsilon }}_1\right)}}{1-\frac{{\left({\mathrm{f}}_1+{\mathrm{f}}_2\right)}^2\left({\mathsf{\epsilon }}_1-{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2+{2\mathsf{\epsilon }}_1\right)+{\mathrm{f}}_2\left({2\mathsf{\epsilon }}_1+{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2-{\mathsf{\epsilon }}_1\right)}{\left({\mathrm{f}}_1+{\mathrm{f}}_2\right)\left({\mathsf{\epsilon }}_1+2{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2+{2\mathsf{\epsilon }}_1\right)+{2\mathrm{f}}_2\left({2\mathsf{\epsilon }}_1-{\mathsf{\epsilon }}_{\mathrm{e}}\right)\left({\mathsf{\epsilon }}_2-{\mathsf{\epsilon }}_1\right)}}$$

(2)

where ε is the average permittivity of the nucleus, ε₁ and ε₂ are the permittivity of ice and dust, f₁ and f₂ are the volume fractions of ice and dust, and ε_e is the vacuum permittivity. From the above, ε is 1.27 ± 0.05 and the volume ratio of dust/ice is between 0.4 and 2.6, from which it follows that the permittivity of the dust will be between 2.9 and 5.4 [56].

The researchers analyzed the CONSERT results for compositional comparisons with materials in the laboratory, including silicate dust and carbonaceous materials. According to the CONSERT permittivity results, a cometary nucleus will have a porosity of 72–87%, 6–12% ice, and 16–21% refractory dust, with carbonaceous material accounting for at least 75% of the volume percent of the dust [56,57].

These data provide important information about the internal structure and composition of comet nuclei and contribute to a better understanding of the process of comet formation and evolution. The Rosetta mission ended in September 2016 with a crash landing on comet 67P. It was the first mission to successfully land on the surface of a comet and the first to use SB-GPR on a comet to study its internal structure.

3.3. Castalia Mission

The Castalia mission, led by the ESA and scheduled for launch in 2028, will target the 133P/Elst–Pizarro main belt comet (MBC) for exploration. MBCs are a newly discovered class of celestial bodies that orbit like asteroids but look like comets [58]. The scientific objectives for Castalia are as follows: (1) to explore the physical mechanisms responsible for the activity of MBCs; (2) to search for water or ice on the surface and in the interior and study the hydrogen–deuterium ratio to test whether MBCs are one of the possible sources of water on Earth; (3) to use MBCs as tracers of the formation and evolution of planetary systems. The flow of the Castalia mission is shown in Figure 8, and it is expected to take approximately 8 years to complete. The Castalia spacecraft will arrive at 133P in 2035 and then surround and explore it for about a year. At the end of the mission, it will attempt to land on the comet. The Castalia spacecraft will carry two types of SB-GPR for the shallow and interior exploration of the comet [45].

3.3.1. SOURCE

SOURCE is a low-frequency SAR which uses Chirp with a center frequency of 20 MHz and a bandwidth of 10 MHz. The SOURCE antenna consists of two 5 m orthogonal dipoles (see Table 3 for detailed parameters). SOURCE is essentially the same design as the SHARAD radar on the Mars Reconnaissance Orbiter [59].

The scientific objective of SOURCE is to derive structural information about the water or ice inside the comet from the time delay and amplitude of the echoes. The team hypothesizes that 133P’s cometary soil is mainly a mixture of ice and rock with a permittivity between 4 and 6. Under these conditions, the maximum depth resolution that SOURCE can achieve is 6 m.

3.3.2. SSR

SSR is a high-frequency SAR using stepped-frequency signals that consists of an antenna unit and an electronic box (Figure 9). The electronics box provides a range of adjustable frequencies from 300 to 800 MHz in standard mode, and up to 3 GHz in extended mode. In standard mode, the SSR can achieve a maximum penetration depth of 50 m and a maximum 3D resolution of 1 m. Measurements in different orbits will allow the realization of three-dimensional 133P stratigraphic imaging and the study of changes in the composition of shallow surface materials. In extended mode, the SSR is capable of using the full bandwidth from 300 MHz to 3 GHz and can achieve a resolution of 6 cm. In this mode, the SSR can calculate the distance between the spacecraft and the comet surface to assist the spacecraft’s attitude and orbit control system in auto-hover mode [45]. The SSR antennas are four orthogonal Vivaldi antennas that transmit a circularly polarized signal and receive two perpendicularly polarized signals (see Table 3 for detailed parameters).

The scientific objective of the SSR is to perform the tomographic imaging of 133P to characterize its internal three-dimensional structure, including stratification, horizontal variability, pits or large buried cavities, and near-surface ice patches [60].

3.4. Tianwen-2 Mission

The Tianwen-2 mission is a long CNSA-led asteroid sample return and MBC surrounding mission as part of the Planetary Exploration of China. Tianwen-2 targets the near-Earth asteroid 2016 HO₃ (469219 Kamo’oalewa), with a diameter of about 40–100 m, and MBC 311P/PANSTARRS, with a diameter of about 320–585 m [12]. Asteroid 2016 HO₃ became a quasi-satellite of the Earth 100 years ago and will remain in this orbit for the next 300 years [61,62].

The flow of the Tianwen-2 mission is shown in Figure 10. Currently, the Tianwen-2 mission is in an advanced stage of preparation. The scientific objective of Tianwen-2 is to surround and explore 2016 HO₃ and collect samples for return to Earth, and then to surround and explore 311P. Tianwen-2 will be launched in 2025. In 2026–2027, it will perform a surround exploration of 2016 HO₃ and then landing and sample collection. In 2027, it will return to Earth to release samples, then travel through Earth’s and Mars’ gravities to reach 311P in 2033, and surround and explore 311P in 2033–2034.

There are two typical types of sampling: anchor-and-attach architecture and touch-and-go architecture. The anchor-and-attach architecture requires complex thrust and attachment systems and is extremely sensitive to surface debris characteristics. However, it can be operated for long periods of time, making sampling more controllable. Completed sample return missions such as OSIRIS-Rex and Hayabusa 2 have used a touch-and-go architecture. This does not require landing and deployment, and can be accomplished in a short period of time in contact with a surface. But, it is more complex for navigation, guidance, and control. For the Tianwen-2 mission, both architectures can be used to ensure that at least one of them works. It should be noted that the anchor-and-attach architecture has no successful precedent [12].

The Tianwen-2 spacecraft consists of an orbiter and a return capsule. The orbiter will be equipped with a two-channel ISAR named ACSR. As shown in Figure 11, it consists of an electronics box and two sets of antennas [63].

Due to the low gravity of the asteroid, the ACSR will fly with 2016 HO₃, creating a relative motion to the radar through the rotation of 2016 HO₃, and thus creating an inverse synthetic aperture effect to obtain echoes [63].

The low-frequency channel of the ACSR has a center frequency of 150 MHz and a bandwidth of 40 MHz. The low-frequency antenna comprises four collapsible 1.2 m dipoles using stepped-frequency signals that transmit a circularly polarized signal and receive two perpendicularly polarized signals. The maximum penetration depth of the low-frequency antenna is 50 m for asteroid targets and 300 m for comet targets, with a vertical resolution of 5 m (see Table 3 for detailed parameters) [63]. The scientific objectives for the low-frequency channel are as follows: (1) to explore and image the internal structure of 2016 HO₃ to determine whether it is a rubble pile or monolithic structure, and to calculate the internal permittivity profile; (2) to image the interior of 311P, search for deep water or ice, make direct observations of the deep structure and material properties, and calculate its internal permittivity profile; (3) to search for internal boulders and layered structures and to determine the circular polarization ratio in order to study the surface roughness and internal rock distribution.

The high-frequency channel of the ACSR has a center frequency of 900 MHz and a bandwidth of 1.2 GHz. The high-frequency antenna is a Vivaldi antenna, which also uses stepped frequency signals. The maximum penetration depth of the high-frequency antenna is 5 m for asteroid targets and 30 m for comet targets, with a vertical resolution of 0.25 m (see Table 3 for detailed parameters) [63]. The scientific objectives for the high-frequency channel are as follows: (1) to image and study the structure and material properties of the shallow surface layer of 2016 HO₃, and to calculate the shallow permittivity profile; (2) to image the shallow layer of 311P, to make refined observations of the structure and properties of the shallow material, and to calculate the distribution of the shallow permittivity profile; (3) to distinguish between shallow surface structures and surface rocks.

ACSR is an important direction for our future work. After ACSR receives the echoes, we will perform signal processing. So, the following will introduce our latest research based on ACSR, and in order to verify whether the imaging algorithm can discriminate asteroids on different surfaces, we have carried out simulation work. Simulation is an important method for radar signal processing. For missions where real signals have been received, simulation can help interpret the signals. For planned missions, simulation can also be used to make a preliminary assessment of the effectiveness of a signal processing method. Assuming that the flight radius of the ACSR is R_g, the flight altitude is Z_h, the azimuth angle is φ ∈ [0, 2π), the coordinates corresponding to each azimuth angle are (R_gcosφ, R_gsinφ, Z_h), the target altitude function is z(x, y), the corresponding three-dimensional distribution function is f(x, y, z(x, y)), and the transmitted signal is p(t), then the echo is

$$\mathrm{s}\left(\mathrm{t},\mathsf{\phi }\right)=\iint \mathrm{f}\left(\mathrm{x},\mathrm{y},\mathrm{z}\left(\mathrm{x},\mathrm{y}\right)\right)\mathrm{p}[\mathrm{t}-\frac{2\sqrt{{(\mathrm{x}-{\mathrm{R}}_{\mathrm{g}}\cos \mathsf{\phi })}^2{+(\mathrm{y}-{\mathrm{R}}_{\mathrm{g}}\sin \mathsf{\phi })}^2+{\left(\mathrm{z}\right(\mathrm{x},\mathrm{y})-{\mathrm{Z}}_{\mathrm{h}})}^2}}{\mathrm{c}}]\mathrm{d}\mathrm{x}\mathrm{d}\mathrm{y}$$

(3)

The Back Projection (BP) algorithm is an algorithm that can be adapted to the needs of surround exploration. The steps are as follows: (1) Grid the imaging area of the target into multiple pixel units and label the index number of each pixel unit in the x, y, and z directions. (2) Consider an echo at an azimuthal angle. The echo with a unit-angle reflector located at the central reference point is matched-filtered in the frequency domain, and then inverted Fourier-transformed to the time domain and phase-compensated. (3) The phase-compensated echoes from all angles are summed to obtain the final image of the target.

The resulting image expression is

$$\mathrm{I}\left({\mathrm{x}}_{\mathrm{i}},{\mathrm{y}}_{\mathrm{j}},\mathrm{z}\left({\mathrm{x}}_{\mathrm{i}},{\mathrm{y}}_{\mathrm{j}}\right)\right)=\mathrm{s}\left({\mathrm{t}}_{\mathrm{i}\mathrm{j}}\left(\mathsf{\phi }\right)-{\mathrm{t}}_0\right)\int \exp \left[\mathrm{j}{\mathsf{\omega }}_{\mathrm{c}}\left({\mathrm{t}}_{\mathrm{i}\mathrm{j}}\left(\mathsf{\phi }\right)-{\mathrm{t}}_0\right)\right]\mathrm{d}\mathsf{\phi }$$

(4)

where

$${\mathrm{t}}_{\mathrm{i}\mathrm{j}}\left(\mathsf{\phi }\right)=\frac{2\sqrt{{({\mathrm{x}}_{\mathrm{i}}-{\mathrm{R}}_{\mathrm{g}}\cos \mathsf{\phi })}^2{+({\mathrm{y}}_{\mathrm{j}}-{\mathrm{R}}_{\mathrm{g}}\sin \mathsf{\phi })}^2+{(\mathrm{z}\left({\mathrm{x}}_{\mathrm{i}},{\mathrm{y}}_{\mathrm{j}}\right)-{\mathrm{Z}}_{\mathrm{h}})}^2}}{\mathrm{c}}$$

(5)

We constructed an asteroid model with low surface complexity (Figure 12a) and an asteroid model with high surface complexity (Figure 12b) based on the open-source software Blender4.1. The dimensions of the model match those of the 2016 HO₃ observations, with a minimum diameter of 40 m and a maximum diameter of 100 m. As shown in Figure 13, the radar is observed in a surround mode with a radius of 600 m.

The radar simulation imaging results are shown in Figure 14 (see Table 3 for the ACSR low-frequency channel for the radar simulation parameters). Figure 14a,b show the first-channel data-matched filter images of the two asteroid model echoes. Both images show peaks at about 500 m from the antenna, the difference being that the low surface complexity has a single peak, while the high surface complexity has two peaks present and branching. As shown in Figure 14c–f, this difference is reflected in the all-channel data-matched filter images, as well as the BP algorithm images.

The reason for this is that the BP algorithm divides the asteroid model into multiple pixel cells, and the echo from a pixel cell can be considered as part of the valid echo only if the vector of the pixel cell’s connection to the radar and the pixel cell’s own normal vector are within the radar beamwidth. As the surface complexity increases, more pixel cell echoes are recognized as valid echoes, thus revealing complex structures in the radar image. Therefore, BP algorithm imaging can characterize asteroids with different surface complexity. We will also continue our work on internal structure analysis and permittivity calculations for asteroid targets based on this characterization.

4. Discussion

The exploration of asteroids and comets has evolved from the first phase of fly-bys to the second phase of surrounding and landing, to the third phase of sample return, which can be summarized as a process from distance to proximity. Future missions will focus more on exploring targets in the second and third phases. At the same time, the missions of the second and third phases seek to obtain information about the internal structure of their targets. As a payload with the advantages of non-destructive, penetrating, and polarization characteristics, SB-GPR has been widely used in lunar and Mars exploration and has been carried on four of the five planned missions for asteroid and comet exploration.

The development of GPR has also gone through a similar process of moving from far to near targets. GB-GPR for long-range observations has evolved into SB-GPR for close-range observations. In this study, statistics on asteroid and comet exploration missions, scientific results, and space-based ground-penetrating radar (SB-GPR) utilization are presented for the three phases to date. According to these statistics, SB-GPR will play an important role in future Phase 2 and 3 missions. The focus of this study was on analyzing the mission flow, SB-GPR parameters, scientific objectives, and scientific results of the missions that have carried SB-GPR and those that are planning to carry SB-GPR, including the Hera, Rosetta, Castalia, and Tianwen-2 missions. On this basis, we believe that the design and signal interpretation of SB-GPR for future asteroid and comet exploration missions will have the following four trends:

(1)

Use of Frequency Modulation (FM) signals: The frequency resolution of the pulsed signals used in conventional GPR is limited by the width of the pulse. The FM signal has a wide bandwidth and is compressed at the receiver end to achieve shorter pulse widths, resulting in higher resolution while maintaining penetration depth. Commonly used FM signals include BPSK-modulated signals, chirp signals, and step-frequency signals.

(2)

Use of multiple frequency bands: Low-frequency signals have a greater depth of penetration but lower resolution, while high-frequency signals have higher resolution but shallower penetration depth. Using multi-band signals to complement each other helps researchers to analyze the surface and internal structure of targets.

(3)

Use of multiple polarization channels: One of the major scientific objectives of asteroid and comet exploration missions is the search for water or ice. Unlike ordinary media, water and ice respond differently to electromagnetic waves of different polarizations, which makes the use of different polarization signals useful in determining the distribution of water and ice on a target.

(4)

Use of multiple explorations for synergistic signal analysis: A common synergistic approach is to use optical payload data in conjunction with radar data. Optical payloads are commonly used to construct three-dimensional models of asteroids or comets and play an important role in radar simulation efforts. Simulation is one of the important methods of radar signal processing, and for missions where real signals have been obtained, simulation can help interpret the signals. For example, the Rosetta mission determined the real part of the average permittivity of the nucleus to be 1.27 ± 0.05 and narrowed the landing area of Philae to 21 × 34 m² by simulation. For planned missions, such as the Tianwen-2 mission, simulation can also be used to make a preliminary assessment of the effectiveness of a signal processing method. For example, in this study, two surface complexity asteroid models are constructed for the ACSR radar of the Tianwen-2 mission, and radar simulation studies are performed. The results confirm that the BP algorithm is capable of imaging asteroid models with different surface complexity, which lays the foundation for our subsequent work on internal structure analysis and dielectric constant calculation.

5. Conclusions

SB-GPR is developing rapidly in asteroid and comet exploration missions, with a focus on the development of FM signals, multiple frequency bands, multiple polarization channels, and multiple explorations for synergistic signal analysis in the second and third phases of missions in the future. It will provide more valuable information for understanding the structure, constituent materials, and evolutionary history of asteroids and comets.