1 Introduction

The current state of understanding of the Moon began with the missions of the 1960s (Rangers, Surveyors, and Lunar Orbiters) as well as Earth-based telescopic studies, that led to the remote sensing, surface exploration, and sample return activities of the Apollo program. Data from more recent lunar missions, in particular the Clementine and Lunar Prospector, and the analysis of newly discovered meteorites of inferred lunar origin, provided further insights into the composition and morphology of the Moon’s surface. Today we are at the dawn of a new era of lunar exploration as demonstrated by the great number of recent, on going and planned missions to the Moon (e.g. SELENE-Kaguya, Chang-e and Chandrayaan seires, Lunar Reconnaissance Orbiter-LRO). In this context we proposed a new Lunar mission, named MAGIA (Missione Altimetrica Gravimetrica GeochImica lunAre), with three main goals: (1) to study the Lunar internal structure and gravity, (2) to perform a detailed morphological and mineralogical mapping of its polar regions, (3) to analyse the lunar exosphere and radioactive environment. All these objectives fully meet the points prioritized by the National Research council for near future lunar exploration (between 2010 and 2020) (Committee on the Scientific Context for Exploration of the Moon 2007). MAGIA will include three main instruments: a high resolution camera (CARISMA), an optical VIS camera and VNIR imaging-spectrometer (CAM-SIR), and a Radar Altimeter (RASCAL). The MAGIA mission strategy is particularly focused on polar regions, where a full resolution coverage is planned, whereas Lunar nearside and farside will be mapped with moderate resolution, except for particular targets where high resolution data are also considered.

In this paper we will illustrate the major geological topics that in our opinion future Lunar missions will have to consider, particularly focussing on crustal and mantle heterogeneities, cold-traps and water deposits, duration and styles of volcanic activity, basin and global tectonics, impact flux and crater chronology. For each topic, we specifically delineate the possible contribution of MAGIA, with a special emphasis on the relation to data acquired by recent and on-going lunar missions.

2 MAGIA Instrument for Geological Interpretation

Among the group of scientific instruments included in the MAGIA payload, the most important ones for geological purposes are the infrared spectrometer and visible wavelength camera called CAM_SIR (CAMera Spettrale InfraRossa), the high resolution visible wavelength camera CARISMA (Camera ad Alta RIsoluzione Spaziale per MAgia), and the radar altimeter RASCAL (Radiometro, Altimetro e Scatterometro Lunare) (Table 1).

Table 1 Technical characteristics and main scientific goals of CAM_SIR, CARISMA and RASCAL instruments
Full size table

The CAM_SIR includes two channels, each one forming a distinct instrument: a Visible (VIS) Imager Channel and a Visible-Near InfraRed (VNIR) Spectrometer Channel (Table 1) (Coradini et al. submitted). The VIS imager is equipped with three filters, ranging from 0.4 to 1 μm, for a low resolution color imaging of the Moon surface. The VNIR imaging spectrometer channel has a spectral range from 0.4 to 5.0 μm and a spectral sampling of 7 nm. Both channels are characterized by a spatial resolution of 23.4 m/pixel at 100 km of nominal acquisition altitude. The main scientific goals of CAM_SIR are the study and mapping of the mineralogical composition of the lunar surface in order to gain new information about the crustal evolution of the Moon. Particular attention will be devoted to distinguish any compositional difference among highlands and mare basalts and to detect signs of any volcanic activity that may have occurred in recent times. Another important aim of CAM_SIR is to detect the possible presence of ice. The comparison with the hyperspectral and multispectral imagers mounted on SELENE-Kaguya, LRO, Chandrayaan-1 and Chang-e spacecrafts (Table 1) highlights the great potential of CAM_SIR in terms of spectral range, spectral resolution and spatial resolution.

CARISMA is a high resolution camera, operating in the visible wavelength range. It includes a panchromatic filter, centered at 650 nm with 500 nm bandwidth, and three broad-band filters (440 nm, 550 nm, 740 nm). Its spatial resolution of 1.25 m/pixel (at 100 km) should identify the surface roughness of the target areas, allow a better detection of geological unit boundaries and, finally, provide key information for recommending future landing sites. Although CARISMA is not particularly innovative in term of spatial resolution with respect to LROC of the Lunar Reconnaissance Orbiter (Table 1), its data acquisition will be complementary to the LROC datasets that are not planned to cover the entire lunar surface (Chin et al. 2007).

RASCAL is a microwave instrument, operating as a radar altimeter, a radiometer and a scatterometer (Table 1) (Orosei et al. submitted). In altimeter mode, the instrument measures the distance between the spacecraft and the surface by emitting a radar pulse and then recording its reflection. Moreover, by analyzing the power and waveform of the radar echo, it is possible to achieve constraints on the dielectric constant and roughness of the surface. In radiometer mode, RASCAL measures the power of thermal emissions in the microwave region in order to derive the brightness temperature, whereas in scatterometer mode, the radar measures the energy backscattered by the observed surface. Both upshots depend on the dielectric constant and roughness of the surface. These data will be acquired according to the following technical requirements: (1) a ground spatial resolution better than 600 m; (2) a height accuracy between 1 and 10 m; (3) a radiometer accuracy better than 1 K and a radiometer resolution better than 0.2 K. The three different modes of operation will be integrated in order to define the Moon’s shape, measure the distance of the surface from the mass centre, produce topographic maps with altimetric observations, measure the surface temperature with radiometric data, analyze the permanently shadowed areas near the Poles and possibly detect water ice within these regions. RASCAL is unique among existing and planned altimeters, which are normally based on LASER technology.

3 Crustal Composition

The Moon is a differentiated planetary body with a crust having an average thickness of 50 km at the near side, 90 km at the farside, and a present lithosphere approximately 1,000 km-thick (Khan et al. 2000; Wieczorek et al. 2006). The surface of the Moon can be divided into two major regions: (a) the older densely cratered and rugged highlands and (b) the younger, smooth, dark areas called maria (17% of the Moon’s surface). The former are mainly composed of Al-rich anorthosites whereas the latter are composed of Fe-rich basalts. Furthermore, three principle groups of rock-types are present in the highlands, as inferred by the analysis of lunar samples and meteorites: the Ferroan anorthosites characterized by a Ca-rich plagioclase (reflecting the low concentration of Na in the Moon), the Mg-rich series of plutonic rocks such as troctolites, norites and dunites, and finally the alkali-suite which includes anorthosites, gabbo-norites and potassium-rare elements-phosporus rich (KREEP) basalts and monzogabbros (e.g. Warner et al. 1976, Papike et al. 1998; Lucey et al. 2006). The maria rocks can be classified on the basis of their TiO2 abundance, in a way that high-Ti (>9 wt% TiO2), low-Ti (1.5–9 wt% TiO2) and very low-Ti (<1.5 wt% TiO2) basalts can be distinguished and in turn subdivided according to their K content (e.g. Neal and Taylor 1992; Papike et al. 1998; Lucey et al. 2006). The mineral assemblages, together with Fe and Ti abundances inferred from remote sensing data (e.g. Clementine imaging spectrometer, Lunar Prospector gamma ray and neutron spectrometers, Chandrayaan-1 M3 imaging spectrometer, LRO-DIVINER radiometer, SELENE-Kaguya Multiband imager) support the presence of Ferroan anorthosites and Mg-rich intrusions in the highlands and basalt with variable Ti content within the maria (e.g. Lucey et al. 2000; Lawrence et al. 2002; Prettyman et al. 2002; Ohtake et al. 2009; Greenhagen et al. 2010; Klima et al. 2010). Indeed, the rugged highlands are characterized by abundant plagioclase (75% on average), which is inversely correlated with the iron content, and minor orthopyroxenes and olivines. In addition, olivine and pyroxene rich locations, possibly Mg-suite rocks, were revealed in several central peaks punctuating the lunar highlands (Pieters 1982; Tompkins and Pieters 1999). On the other hand, maria show the typical mineral-assemblages of basalts with clinopyroxenes (positively correlated with Fe) and olivines.

3.1 Magma Ocean Hypothesis

The rock variety and related surface distribution on the Moon seem to find its explanation in the simple differentiation of a primordial magma ocean which led to a typical global layering (Fig. 1a) (Warren 1985). Partial melting of such a primordial layering should have had influenced the composition of the later basalts infilling maria basins.

Fig. 1
figure 1

Magma ocean Hypothesis: a Mantle layering and effusion sequence (after Shearer et al. 2006), b Mg and REE abundances with depth (after Shearer et al. 2006), c REE patterns in Ferroan anorthosites, KREEPs and maria basalts (after Taylor 2001)

Full size image

According to most models, the crystallization sequence would have been Ol → Opx + Ol → Ol + Cpx + Opx + Pl  → Cpx + Pl  → Cpx + Pl + Il (Taylor 1982; Shearer and Papike 1999; Shearer et al. 2006). Hence, after a first stage of sinking of olivines and orthopyroxenes (to form the deepest mantle cumulates), the lighter plagioclase started to float and gradually create the primordial crust. The crustal formation was accompanied by new mantle cumulates with increasing ilmenite content. The final stage was the crystallization of residual melts enriched in incompatible elements, the urKREEPs sandwiched between the ferroan anorthositic crust and the uppermost mantle cumulates.

The strongest support for the magma ocean hypothesis is the incompatible element content of lunar rocks. An increase in U, Th, K, REE is indeed shown from the Ferroan anorthosites to the “KREEPy” rocks, whereas a strong depletion of the same elements are recorded on mare basalts that should reflect the composition of their deep mantle sources (Fig. 1b). This effect is coupled with the smooth transition between the Europium rich ferroan anorthosites and the Europium-depleted KREEPs and maria basalts (Fig. 1c) (e.g. Taylor 2001).

The magma ocean model and consequent mantle layering implies a specific evolution of the Lunar magmatism that would have reflected the progressive deepening of the magmatic sources. Hence, the Mg-suite rocks should have followed the formation of the Ferroan anorthositic crust and possibly recorded a transition from the magma ocean to the serial magmatism era (e.g. Longhi 1980; Shearer and Newsom 2000; Shearer et al. 2006). The Al suite intrusions represent plutons well inside the serial magmatism stage. As far as effusions are concerned, the former KREEP basalts were followed by maria effusions with an expected Ti content decreasing through time (Fig. 1a) (Taylor 1982).

3.2 A Heterogeneous Crust

Although there is overwhelming evidence that the broad picture of the magma ocean is substantially correct, it fails to take into account several indications of a heterogeneous composition of both the Lunar crust and mantle.

The first challenge to a simplified magma ocean hypothesis is the age of the lunar rock families. Although KREEP (3.8–4.0 Gyr) and maria basalts (3.9–1.1 Gyr) are indeed younger than the ferroan anorthositic and Mg-suite rocks, there is a considerable overlap between the ages of these latter groups (4.5–4.3 Gyr and 4.43–4.17 Gyr, respectively; Taylor et al. 1993; Borg et al. 1999). This means that Mg-suites should have played a substantial role even in lunar primary crust formation, which should have been much more complex than generally supposed (e.g. Shearer et al. 2006). Further evidence is provided by the widespread presence of mafic-rich materials in the lowermost layer of the crust as suggested by geophysical observations and modelling (Khan et al. 2000, Wieczorek and Zuber 2001; Khan and Mosegaard 2002), spectral evidence of norites, troctolites and dunites in basin rings and crater peaks (Tompkins and Pieters 1999; Hawke et al. 2003) and the possible noritic composition of the South Polar Aitken basin floor (Lucey et al. 1995; Pieters et al. 1997, 2001).

The formerly hypothesized progressive decrease of TiO2 in maria basalt (Taylor 1982, Haskin et al. 2000) has been seriously disputed by several authors (Hiesinger et al. 2000, 2003; Hackwill et al. 2006, Hackwill 2010) who point to the absence of any obvious correlation between age and FeO and TiO2 abundances in these rocks; in addition Bugiolacchi and co-workers (Bugiolacchi et al. 2006; Bugiolacchi and Guest 2008) demonstrated an overall, but substantial, increase of TiO2 and FeO content during Imbrium mare emplacement. This implies heterogeneous magmatic sources throughout the lunar mantle and/or different regimes of magma transport and crystallization.

Gravitational instabilities and mantle overturning phenomena, due to the late formation of dense ilmenite-bearing cumulates in the crystallizing magma ocean, have been advocated to explain such mantle heterogeneities, but a general consensus on this subject is still far off (e.g. Elkins-Tanton et al. 2002; Longhi 2003; Stegman et al. 2003; Shearer et al. 2006).

The most obvious indications of a complex and heterogeneous lunar crust are the Th and Sm maps from Lunar Prospector gamma ray spectrometer (Elphic et al. 2000; Jolliff et al. 2000; Lawrence et al. 2000, 2007). These maps show a dramatically uneven distribution of REE-elements (hence KREEP rocks) that are strongly focussed around the ramparts of the Imbrium impact basin and virtually absent within the huge South Polar Aitken basin where lower-crust or even upper mantle materials are expected to outcrop in places. The geochemically anomalous areas of Imbrium, called the Procellarum KREEP Terrane (Jolliff et al. 2000), have been explained by the effects of the two superimposed Procellarum and Imbrium impacts that excavated the KREEP layer at the lowermost crust-upper mantle transition (e.g. Hiesinger and Head 2006). However the virtual absence of important REE-rich areas in the deep South Polar Aitken basin points, once more, to a strongly heterogeneous lunar deep-crust and mantle. In this view, even the step shown by the seismic velocity profile at 500 km underneath the near side maria (Khan et al. 2000) may be a local effect reflecting a mantle heterogeneity, rather than a global feature. This can be explained as a result of the near-side magmatism that caused a very localized Alluminium depletion in the upper mantle (e.g. Wieczorek et al. 2006).

Olivine, spinel and orthopyroxene rich materials detected throughout the lunar surface by the spectrometers aboard the Clementine, SELENE-Kaguya and Chandraayan-1 missions (e.g. Pieters and Tompkins 1999; Pieters et al. 2001; Yamamoto et al. 2010; Pieters et al. 2010) as well as the highly siliceous outcrops very recently discovered by the LRO-DIVINER radiometer (Glotch et al. 2010) further complicate this general framework.

3.3 The Magia Contribution to the Assessment of Crustal Composition

Most of the known petrology of the lunar crust has been derived from rock and soil samples restricted to a very limited region on the equatorial nearside. The data provided by the Clementine and Lunar Prospector, as well as the ones derived by the current missions (Chandaryan-1, Chang-e, SELENE-Kaguya, LRO) have highlighted a very heterogeneous crust, suggesting complex scenarios of crustal evolution. To unravel these, multi-sensor data acquisition is required as well as synoptic geological maps that integrate surface morphology, stratigraphy, tectonics and composition. However, to date remote-sensing data are still of varying quality and, although they have permitted the production of several mineral and elemental maps (e.g. McEwen et al. 1994; Lucey et al. 1994; Elphic et al. 2000; Jolliff et al. 2000; Lawrence et al. 2000, 2007; Chevrel et al. 2002; Hawke et al. 2003; Lucey 2004; Giguere et al. 2006), only some local geological sketches (e.g. Heather and Dunkin 2002, Heather et al. 2003; Hawke et al. 2005; Bugiolacchi and Guest 2008) were derived. A better integration of surface composition and geological processes is needed at this stage and constitutes one of the major objectives of the MAGIA mission.

Due to its spectral and spatial resolution, MAGIA CAM-SIR will be highly complementary to the SELENE-Kaguya, Chang-e and Chandrayan-1 imaging spectrometers and, particularly if integrated with the LRO-DIVINER thermal data, it will provide a reliable mineralogical investigation of the lunar surface (Coradini et al., submitted). Subsequently, these data could be compared to the elemental analyses from X-ray, gamma-ray and neutron spectroscopy, and α-particle backscatter provided by the recent and on-going Lunar missions. The multi-sensor data integration and comparison for composition assessments should represent the basis for further work aiming to relate stratigraphy and tectonics to petrography. In the Polar regions, high resolution MAGIA mission data should provide the quality data necessary to begin this process, by integrating the mineralogical mapping of CAM_SIR, the stratigraphic information provided by CARISMA and the topographic measurements from RASCAL.

4 Water on the Moon

In agreement with the giant impact hypothesis (e.g. Canup 2004), the early Moon lost most of its volatiles during the severe impact heating and the resulting formation of a wide magma ocean. This seems to be suggested by the lack of indigenous water and OH− ions in the samples acquired during the Luna and Apollo programs (e.g. Lucey et al. 2006). Nevertheless, water has recently been identified in lunar glasses (Saal et al. 2008) and apatites (Boyce et al. 2010) and the neutron spectrometer data from the Lunar Prospector mission found that hydrogen is present in the permanently shadowed areas within craters of the polar regions. These “cold traps” allow the accumulation and retention of volatile elements coming from external sources such as impacted comets and water-bearing asteroids (Watson et al. 1961; Nozette et al. 1996; Feldman et al. 2000; Crider and Vondrak 2002; Bussey et al. 2003).

Recently the Moon Mineralogy Mapper spectrometer (M3) of the first Indian lunar mission Chandrayaan observed that hydration processes do occur on most of the satellite surface by detecting absorption features near 2.8–3.0 μm, that are highly diagnostic of OH− and H2O (Pieters et al. 2009b). The M3 global spectral pattern shows an increase of the absorption bands with latitude and no absorptions were detected in the equatorial regions. This is possibly due to the masking effect of a minor thermal emission component beyond 2.6 μm at lowermost latitudes (Clark 1979). Indeed a revisited interpretation of data collected by the Cassini Visual and Infrared Mapping Spectrometer (VIMS) and Deep Impact High Resolution Instrument—infrared spectrometer (HRI-IR) has concluded that the entire lunar surface is hydrated for at least some portions of the lunar day. This suggests that instantaneous temperature due to solar radiation is the dominant factor determining the OH− and H2O abundances recorded by these instruments. The daily hydration and dehydration processes could be stoked by solar wind and lead to a migration of OH− and H+ towards more thermally stable zones such as polar regions (Clark 2009; Sunshine et al. 2009). This mechanism is further supported by the measurements of the neutral mass spectrometer of the Chandrayaan mission (CHACE-Chandra’s Altitudinal Composition Explorer) which found a clear trend of increasing concentration of H2O with latitude (Sridharan et al. 2010).

To verify the presence of water deposits in the south polar region, the permanently shadowed areas of Cabeus crater were hit by the upper stage of the NASA rocket LCROSS (Centaur). The plume due to the impact was analyzed by the ultraviolet/visible and near-infrared spectrometers, revealing the occurrence of OH− ions and implying the existence of water ice on the crater wall (Colaprete et al. 2010). On the other hand, albedo observations from the panchromatic Terrain Camera of SELENE-Kaguya have not recorded water–ice deposits within the south polar Shackleton Crater (Haruyama et al. 2008). In conclusion, in lunar polar regions the volatile content within craters may either vary consistently or outcrop differently and its abundance and physical state are still unknown. Hence a very detailed analysis of these zones—including geological interpretation and absolute age assessment of the deposits—and a theoretical study of the survival of possible ice deposits are emphatically needed.

4.1 The MAGIA Contribution to the Quest for Water

One of the main contributions of MAGIA will be to clarify the morphology of the polar regions. The RASCAL altimeter will obtain precise elevation measurements useful for calibrating the Digital Terrain Models derived from the SELENE-Kaguya and Chang-e stereo-cameras. This detailed topography with reliable absolute elevations will allow the identification of potential cold traps together with accurate estimates of their extension and inferences on the volume of possible water deposits. CARISMA will integrate this information with high resolution topographic and geological details, such as surface extent and thickness of regolith layers that, thanks to their thermal shield effect, may favor the maintenance of the water deposits.

An additional contribution of RASCAL in its scatterometer mode will be to measure the permittivity of target areas and hence infer material porosities and water content, even in shadowed regions.

Finally the VNIR spectrometer of CAM_SIR will infer the presence of hydrogen and its compounds in non-shadowed areas of both the farside and polar regions. Any potential finding could be subsequently compared to the OH− and H+ exosphere abundances revealed by the MAGIA Neutral Particle Monitor ALENA (Analizzatore Lunare di ENA) (Milillo et al. submitted).

5 Volcanism

Lunar volcanism occurred mostly as basalt maria filling the extended basins of the nearside (Head 1976; Head and Wilson 1992; Hiesinger et al. 2003) and a limited number of basins and craters of the lunar farside and polar regions (Haruyama et al. 2009; Morota et al. 2009).

Most of the magmatism mainly occurred during early lunar history from about 3.9 Ga to 3 Ga (Wilhelms et al. 1987), when the high frequency of large impacts, coupled with higher geothermal gradients, may have even caused impact induced volcanism (Ivanov and Melosh 2003 ). However, mare volcanism continued up to very recent times, since some nearside regions show an age of 1.1 Ga (Hiesinger et al. 2000), whereas farside basalts may reach ages as young as 2.2 Ga (Haruyama et al. 2009).

Thanks to the large amount of data from remote sensing acquisitions and Apollo samples, the nearside maria are better characterized, from the morphological and mineralogical point of view, than farside and polar maria. Indeed, these latter are still largely unknown, although very recently the SELENE-Kaguya data allowed a more accurate study of their geological aspects (Haruyama et al. 2009; Morota et al. 2009). Finally, several still poorly known magmatic clues were detected within the South Pole-Aitken basin. The instruments of MAGIA, in particular CAM_SIR, will provide important contributions to the maria characterization, especially concerning those located on the farside and poles. Moreover these cameras will allow us to study in more detail the numerous features associated with lunar maria, such as lava flows, volcanic domes and dark mantling deposits (e.g. Milkovich et al. 2002).

5.1 Lava Flows and Effusion Rates

Lunar lava flows may reach lengths of 100 km (Wilhelms and McCauley 1971; Zimbelman 1998). Indeed, the low gravity of the Moon implies a minor buoyancy of magmas within the crust, and consequently effusions can be accomplished by great volumes of ascending magmas, thus implying greater effusion rates through larger conduits (106–109 kg s−1 vs 102–107 kg s−1 for basaltic effusion in the Earth, Whitford-Stark 1982; Wilson and Head 1983). This partly accounts for the difference in length between lunar and terrestrial flows (250–1,000 km vs maximum 180 km of sub-aerial flows, respectively). However, studies of basaltic lava flows in Hawaii (e.g. Walker 1991; Hon et al. 1994) suggest that a thermally efficient mechanism of lava flow spreading such as inflation (Zimbelman 1998) or tube feeding (Keszthelyi 1995; Dragoni et al. 1995; Sakimoto et al. 1997) should also be considered to explain the long lunar flows. Indirect estimates of lava flow and lava units thickness on the Moon are extremely variable (20–60 m for the near side; Howard et al. 1972; Schaber 1973; Schaber et al. 1976; Neukum and Horn 1975; Gifford and El-Baz 1981; Hiesinger et al. 2002), suggesting that different emplacement processes (normal or inflated flows) are likely to be present on the Moon. Within this framework stands the long lasting debate about the origin of the sinuous rilles that can be interpreted both as lava channels, formed through thermal and mechanical erosion of the substrate by low viscosity lavas (Hulme 1973; Carr 1974; Head and Wilson 1981; Williams et al. 2000), and/or collapsed lava tubes (Oberbeck et al. 1969; Greeley 1971; Spudis et al. 1988). Recently, the high resolution Kaguya-SELENE Terrain Camera images allowed Haruyama et al. (2009) to detect a skylight, located along a sinuous rille of the Marius Hills region (Fig. 2). This finding unambiguously proves a potential endogenous spreading mechanism for several lava flows in lunar basins.

Fig. 2
figure 2

Skylight located near the sinuous rille, named Rille A, in the Marius Hills region. It is interpreted as a roof collapse of a lava tube (Haruyama et al. 2009)

Full size image

5.2 Dark Halo Craters and Pyroclastic Activity

Explosive eruptions develop from a magma approaching a planetary surface when the bubble volume fraction in the melt exceeds about ¾, and the pressure at which magma disruption occurs is greater than the atmospheric pressure (Whitford-Stark 1982; Wilson and Head 1983). Therefore on the Moon, where atmospheric pressure is negligible, magma disruption into pyroclasts should have been a quite common event as proven by deposits of likely pyroclastic origin occurring at lunar maria margins, both on plains and uplands (Head 1974; Head and Wilson 1979). The pyroclastic origin of these units is suggested by their singularly low albedo (Fig. 3a), smooth surfaces and subduing relations with the underlying terrain (Wilhelms and McCauley 1971; Hawke et al. 1989; Gaddis et al. 2010).

Fig. 3
figure 3

Dark mantling deposits. a pyroclastic deposits in the western Aristarchus plateau. The bright-rayed crater at right-center of the image has penetrated the pyroclastic deposits and exposed fresh, light-colored rocks of the lunar highlands (LROC NAC image M111945148R); b several dark halo craters (white arrows) located along floor fractures within the Alphonsus crater (108 km of diameter) (Lunar Orbiter image IV-108-H2)

Full size image

Among dark mantling deposits are dark-halo craters (Fig. 3b), even though their pyroclastic nature has been locally questioned, since dark material within craters may be related either to impact melts (Howard and Wilshire 1975; Hawke and Head 1977, 1979; Hawke et al. 1979), or to dark deposits excavated from beneath (Hawke and Belt 1981). In the latter case the dark halo craters may be indicators of the presence of crypto-maria, which are mare basalt hidden by superposed brighter material (Antonenko et al. 1995).

Although an accurate spectral characterization is essential to distinguish dark-halo craters of impact origin from those formed by volcanic processes, morphology, texture and structures can also aid discrimination between these two kinds of feature (Head and Wilson1979).

5.3 Volcanic Domes

Lunar domes are characterized by a circular to rather irregular outline, a general convex shape, relatively low slopes (generally less than 5°), and diameters ranging up to about 30 km. Some of them show summit craters, and virtually all occur in association with extensive maria plains (Fig. 4) (Head and Gifford 1980). The formation of lava domes is still uncertain, and thought to be related either to slow eruptions of viscous lavas (Head and Gifford 1980), intrusions of shallow laccoliths (e.g. Michaut 2010; Wöhler et al. 2010), or mantling of large blocks with younger lavas (Heather et al. 2003; Lawrence et al. 2005). The correct interpretation of these features is important to understand both the Lunar crustal evolution and magmatism.

Fig. 4
figure 4

Image taken by HDTV of Kaguya showing a volcanic dome field of the Marius Hills region. Most of the domes are characterized by low step flanks and are interpreted as shield volcanoes

Full size image

5.4 The MAGIA Contribution to Understanding Lunar Volcanism

Color-images from CAM_SIR will be useful for recognizing possible compositional variation among the different eruption phases, especially on the South Pole-Aitken basin maria. These data could be integrated with the images of LROC or CARISMA that, with their high spatial resolution, will enable the detection of particular key features, like truncated sinuous rilles, embayed domes and partially buried craters, which could be crucial to establishing the relationships between the different magmatic phases.

MAGIA instruments will also allow a better discrimination among different eruption styles that occurred on the Moon, and provide important new information for the study of lava flows. For example the high vertical accuracy of RASCAL altimeter data, combined with the 3D rendering of SELENE and Chang-e cameras, should better constrain the thickness of the lunar lava flows. Furthermore, the high spatial resolution of CARISMA will considerably enlarge the counts of crater size frequency distributions towards small diameters (CSFDs), essential for deriving the thickness of thin lava flows and volcanic deposits (see for example, Neukum and Horn 1975; Hiesinger et al. 2002, 2003; Platz et al. 2010). This study will improve our knowledge of the effusion rates during lunar evolution. CARISMA also has great potential for discriminating the fingerprints of endogenous lava emplacement (i.e. tumuli, lava rises, lava ridges, lava tubes and skylights) in places that include those where LROC data will be not available (Fig. 5).

Fig. 5
figure 5

Evidence of endogenous spreading mechanisms for lava flows detected on both martian and terrestrial flows. a tumulus identified on a Daedalia Planum (Mars) lava flow surface (PSP_002711_1550); b several lava ridges which surround pre-existing volcanic cones on the Pampas Onduladas lava flow (Argentina) (Pasquarè et al. 2008) and c lava rise detected on Pampas Onduladas lava flow (Argentina) (Pasquarè et al. 2008)

Full size image

Data integration between the high resolution images from LROC and the spectral data of CAM_SIR could shed light on the extent and compositional variations of the lunar pyroclastic deposits. In addition, dark halo craters will be easily detected and classified, especially on the farside and polar regions. This could potentially provide a better knowledge of the extent of crypto-maria as well as of pyroclatic deposits, both extremely important for understanding the extent and style of lunar magmatism. CARISMA and CAM-SIR VIS images, coupled with the 3D data of SELENE (in turn calibrated with the altimetric data of RASCAL), could be very useful for detecting and classify different volcanic forms. A detailed study of volcanic forms should include a multispectral analysis since different morphologies could also be related to different compositions (e.g. Besse et al. 2010). For example, the spectral differences between a given volcanic edifice and the underlying terrain may enable us to distinguish between an endogenous volcanic dome and a cinder cone or a strato-volcano. CAM_SIR VNIR data will provide such information.

Finally MAGIA has the potential to provide a valuable contribution to dating the lunar volcanic units. The high spatial resolution of CARISMA and the synthetic view of CAM-SIR VIS will enable crater counts in a broad range of diameters that will be extremely useful for the age determination of both extended lunar maria and confined lava flows. This, coupled with the compositional information derived from CAM_SIR VNIR products, will provide better constraints on the volcanic and magmatic evolution over the lunar geological time.

6 Tectonics

The thicknesses of the Moon’s crust and lithosphere (50–90 km and 7–1,000 km, respectively) in comparison to its radius (1,738 km) coupled with a low heat flux (0.03 J/m2s vs 0.02–3 J/m2s for the Earth) account for a prevailing conduction heat transfer mechanism for this planetary body. In such a case, tectonic features on the surface do not directly depend on mantle motion and can be generally ascribed to other processes like dynamic loading (impact cratering), contractional cooling, tidal deformations, local and not completely compensated isostatic movements, and very local gravitational and volcano-tectonics processes. In planets dominated by conduction mechanisms, resurfacing events are rare and isolated and the early tectonic history of the planet is preserved for a long time. In addition, a thick lithosphere with a crust characterized by low quartz content is hardly deformable, but once deformed easily retains structural relief preventing topographic relaxation especially on a low gravity planet. This explains the chance to recognize ancient (4.3–3.9 Ga) multiring basins on the Moon such as for instance the Mendel-Rydberg, Mutus-Vlack and South Pole Aitken basins (Wilhelms et al. 1979; Spudis et al. 1994). A prevailing convection associated with a thinner lithosphere can however be envisaged in the very early lunar geological history, as invoked by Lucey et al. 1994 as an explanation of lunar dichotomy alternative to the cratering hypothesis. Nevertheless, a substantial increase in thickness of the elastic lithosphere after maria basin formation and the time when volcanic filling of maria ceased is proven by the mascon gravity anomaly that demonstrates a lithosphere thick enough to support the final maria load (e.g. Head and Solomon 1981).

Whatever the early tectonic history of the Moon might have been, today the main morphological structures are characteristically associated with basins formation and evolution. They can be subdivided into wrinkle ridges and normal faults, often forming grabens, troughs and linear rilles. In addition several lobate scarps similar to the Hermean ones, but with a more limited extension and less relief, have been detected throughout the lunar surface and associated with possible contractional cooling processes.

6.1 Wrinkle Ridge, Grabens and Basins Evolution

A large body of literature has accumulated up to the present day dealing with the study of wrinkle ridges since similar features have also been found on other terrestrial bodies (Mars and Mercury). These features, which can be up to 150 km long, are characterized by (1) a broad rise (25 km wide and 500 m high) (2) a superimposed often asymmetric arch (6 km wide and 200 m high) and (3) a crenulated ridge (200 m wide and 100 m high) (Fig. 6) (Strom 1972; Maxwell et al. 1975; Lucchitta 1977; Golombek et al. 1991; Sharpton and Head 1988; Watters and Johnson 2009). The origin of these structures was widely debated with arguments favouring volcanic and intrusion mechanisms (e.g. Quaide 1965; Strom 1972; Scott 1972; Hodges 1973) in contrast to the tectonic faulting-folding models, the latter particularly supported by the Apollo radar sounder data revealing the maria subsurface setting (e.g. Gilbert 1893; Ronca 1965; Bryan 1973; Howard and Muehlberger 1973; Maxwell et al. 1975; Lucchitta 1976; Maxwell 1978). More recent papers dealing with this argument support a tectonic origin on the basis of ridge-crater intersections (Sharpton and Head 1988), topographic analysis (Golombek et al. 1991), Earth analogues (Plescia and Golombek 1986; Watters 1988) and SELENE-Kaguya sounder data (Ono et al. 2009). Nevertheless doubts remain regarding the different deformation styles that can range from simple buckling (Muehlberger 1974; Sharpton and Head 1988), to high angle faulting (Lucchitta 1976; Sharpton and Head 1988), fault-propagation folding (Golombek et al. 1991; Plescia and Golombek 1986) and low angle fault-bend faulting (Golombek et al. 1991; Plescia and Golombek 1986; Schultz 2000). These uncertainties are all strictly related to a thin vs thick skinned tectonics debate which has a great influence on shortening and offset estimates inside maria, and consequently on the evaluation of their subsidence. Other uncertainties come from wrinkle ridges kinematics, which can be compressive, as suggested by Plescia and Golombek (1986) and Golombek et al. (1991) or even transpressive, as appears to be the case for some en-echelon arrangements (Tjia 1970; Bryan 1973; Fagin et al. 1978; Sharpton and Head 1988). Kinematic interpretations are further complicated by changes of the morphological asymmetry recorded along some important wrinkle ridges (e.g. Dorsum Nicol in Mare Serenitatis, Watters and Johnson 2009). Additional clues about wrinkle ridges are related to the effects of pre-existing topographic heights in the pre-mare basement on ridge localization (e.g. Maxwell et al. 1975; Sharpton and Head 1982; Wilhelms et al. 1987).

Fig. 6
figure 6

Lunar wrinkle ridges detected on Mare Serenitatis (Apollo image AS17-602) (a) and near Montes Teneriffe (LRO NAC image M102264014RE) (b)

Full size image

Despite the extensive works on wrinkle ridges, a detailed analysis of normal faults and grabens is to date very limited. Grabens and linear rills are normally interpreted as concentric extensional features developed along the basin’s edge and neighbouring areas in response to stresses caused by the loading of maria basalts (Fig. 7) (Bryan 1973; Maxwell et al. 1975; Lucchitta 1976; Head and Solomon 1981; Fleitout and Thomas 1982; Wilhelms et al. 1987). Nevertheless, Lucchitta (1976) demonstrated that two scarps (Lee-Lincoln and West Serenitatis) are directly connected to the Mare Serenitatis compressional wrinkle ridges; therefore the kinematics of these features are not yet fully constrained.

Fig. 7
figure 7

Some extensional structures detected on the lunar surface. a a portion of the central section of the Rima Ariadaeus about 120 km in length (AS10-H- 314645); b LROC NAC image showing a sector of Rima Burg, a long graben cutting Lacus Mortis and extending into the adjacent highlands (LRO NAC image M102006420RE)

Full size image

Near-edge basin grabens and inner basin wrinkle ridges seem to be the result of the same phenomena. The one plate shell crustal structure and the rapidly increasing lithospheric thickness of the Moon impede the fully isostatic equilibrium of most of the basins after impact dynamic loading, mantle overcompensation and/or maria basalt infilling. These are the main causes invoked to explain the positive gravity anomalies associated with maria, the so called mascons (e.g. Konopliv et al. 2001). For all basins, the isostatic compensation is facilitated by the extensional structures at the basin edge, which often cannot fully develop due to the lateral constraints. This may cause later thrusting and folding inside the maria themselves, and the consequent formation of wrinkle ridges, possibly enucleated along pre-existing basin structures such as inner rings. Wrinkle ridges could have also been further facilitated by global contraction due to the generalized cooling of the planetary body (Solomon and Head 1980). Such an evolution seems to be established by the relative and absolute model ages of the units deformed by these structures, since normal faults and grabens do not cut maria plains younger than 3.6 Ga (Lucchitta and Watkins 1978), whereas wrinkle ridges deform plains as young as 1.2 Ga (Hiesinger et al. 2003; Watters and Johnson 2009).

Although self consistent specifically for mascons, this general evolution is still not entirely understood for basins not characterized by a positive gravity anomaly; much more surface and subsurface data and modelling are required to cast further light on maria tectonics.

6.2 Lobate Scarps and Global Cooling

If the literature on grabens is still quite poor, the lobate scarps are even less well evaluated since their limited length (mean of 4.4 km) and a very low relief (no more than 80 m) have substantially hindered their detection and consequently the comprehensive mapping of these structures. By analogy with Mercury these structures are interpreted as thrust faults related to planetary cooling and developed in the last phases of lunar geological evolution (Solomon and Head 1979; Binder and Lange 1980; Binder and Gunga 1985; Watters and Johnson 2009). The lobate scarps show indeed a clear post-maria development (Fig. 8), and the analysis of their spatial distribution and contractional displacement led some authors to suppose a planetary radius decrease of 70 m from the time of their formation (ca. 1 Ga, Watters and Johnson 2009). However this estimate is very preliminary due to the limited coverage of high resolution and 3D data for the lunar surface. In addition, if the supposed global cooling is perturbed by other processes, either global such as tidal effects or local such as basin tectonics, the lobate scarp orientations should reflect the related stress field alterations and be less randomized than expected from the cooling contraction alone. To date, a thorough study of lunar lobate scarps has not being carried out and very little can be said about the relationship between tides and/or basin tectonics and global cooling contraction.

Fig. 8
figure 8

A lobate scarp (white arrows) cutting the mare basalt deposits of Aitken Crater, on the lunar far side (LRO NAC image M103374879RE)

Full size image

6.3 The MAGIA Contribution to Lunar Tectonics

In order to improve our knowledge of lunar tectonics, several datasets are still required. Accurate 3D topography from stereo reconstructions, interferometry, radar or laser altimeters will allow the detection and more precise characterization of the morphology of multi-ring basins, wrinkle ridges, lobate scarps and grabens. Radar sounding experiments like the SELENE-Kaguya study can investigate structures in depth, constraining the dips and offsets of wrinkle ridge and graben related faults. High resolution images can detect small-scale structures and give important information about kinematics and cross-cutting relationship among different features, constraining the tectonic evolution throughout Lunar history. Medium-resolution hyperspectral acquisition may allow an easier detection of faulted boundaries between units with different spectral characteristics. To obtain all this specific data at the same resolution for the whole lunar surface numerous missions are required and the data provided should be highly complementary. MAGIA acquisition has been specifically planned for this purpose. For example the 3D rendering from the SELENE-Kaguya terrain camera coupled with high resolution images from CARISMA and LROC and depth measurements from SELENE-Kaguya lunar radar sounding may constrain the presence of faults, their kinematics, dips and even depths. In addition, although not equipped with a stereo camera or a laser altimeter, MAGIA will provide further topographic data through its Radar altimeter device (RASCAL) to be integrated with the Clementine and LRO laser measurements and SELENE-Kaguya and Chang-e stereo and laser altimetry acquisitions.

MAGIA will also obtain useful regional views through its imaging spectrometer CAM-SIR that will facilitate the detection of potential major tectonic features separating different spectral units. In the meantime the high resolution of CARISMA images together with the existing LROC images will probably improve our knowledge of lobate scarp distributions and will also be used to understand their relationship with grabens and, most interestingly, with wrinkle ridges that in several cases may represent their direct prosecution (Lucchitta 1976; Raitala 1984; Watters and Johnson 2009). The particular focus of MAGIA on the polar regions, where mascons are very limited if not totally absent, will give us the opportunity to specifically address features related to global cooling contraction and other possible non-mascon associated structures. Finally, a comprehensive mapping of the structural features at the lunar surface obtained through the integration of data provided by different missions (MAGIA included), will provide important insights into the possible relationships between moonquakes sources, inherited structures and lunar active tectonics.

7 Cratering Record

Impact cratering can be considered the main tectonic process on the lunar surface (e.g. Head and Solomon 1981). Dynamic loading causes: (1) cavity excavation; (2) massive sub-lithospheric flow responsible for dynamic rebound, rim and peak development; (3) significant changes in the physical properties of the lithosphere that can lead to extensive post-impact relaxation, (4) triggering of lava effusion. The analysis of large and small-scale deformation features related to Lunar impact craters, including rings, rims, peaks, faults, fractures and antipodal deformation structures, is essential for the understanding of impact tectonics, whose study is otherwise difficult on the few, heavily degraded or almost completely buried, Earth craters. As already described, craters are also of paramount importance to assess the mineralogical zoning of the upper hundred meters of the crust or even the composition of lunar mantle (e.g. Tompkins and Pieters 1999). In addition, the morphology of craters is also of great value for tectonic studies since craters readily record deformation processes and fault displacements; at the same time, studies of impact processes and Crater Size Frequency Distributions (CSFDs) can shed light on the rheology of the target itself (Holsapple and Housen 2007; Massironi et al. 2009; van der Bogert et al. 2010). Besides these important applications, the cratering record is crucial for assessing the evolution of the impact flux, ages and even thicknesses of geological units on planetary surfaces (e.g. Neukum et al. 2001; Marchi et al. 2009; Hiesinger et al. 2000; Hiesinger et al. 2003; Platz et al. 2010). However, strong debate still persists on the variability of the impact flux during Solar-System evolution, the biases due to the contribution of secondaries to the cratering record, the application of different production functions models to fit CSFDs and the calibration of the lunar chronology. Since MAGIA may substantially contribute to such topics, we will specifically review the relationships among cratering records, lunar chronology and impact flux.

7.1 Crater Chronology

The geological exploration of any planetary bodies has the primary objective of understanding its surface history, which is in turn related to its thermal and tectonic evolution. Clearly, it is of paramount importance to constrain the ages of any geological events recognized on planetary surfaces, by using either relative or absolute chronology. The former is attained by defining the stratigraphical relationships among geological units and cross-cutting interplays between different forms and structures, whereas the latter is obtained through radiometric dating of samples. The Moon is the only planetary body, other than the Earth, for which both chronologies have been successfully applied. Moreover, the extremely well preserved cratering record of its surface permitted the development of a specific chronology based on the assumption that craters form randomly and were accumulated through time at known average rates as a function of size. On this basis, both relative (e.g. Strom et al. 2005) and absolute models of crater chronology (e.g. Hartmann et al. 1981; Neukum and Ivanov 1994; Marchi et al. 2009) were achieved for the Moon and then exported to other planetary bodies of the Solar System. At present two main methods have been developed in order to obtain an absolute chronology model from crater counts. The first one assumes that the lunar CSFDs are characterized by a constant shape over the past 4 Gyr and can be fit by a polynomial of 11th degree known as Neukum Production function (NPF, Neukum 1983; Neukum and Ivanov 1994; Neukum et al. 2001). The second method is based instead on the impactor size and velocity distributions, derived from dynamical models describing the evolution of Main Belt Asteroids (MBAs) and Near Earth Objects (NEOs) (Bottke et al. 2002, 2005; Marchi et al. 2005), and then converted into the crater Model Production Function (MPF; Marchi et al. 2009) through the Holsapple and Housen (2007) scaling law. This latter procedure may take into account a variable flux of impactors through time, as well as different rheological layering of the target. However, both the NPF and MPF models depend on the evolution of the impact flux, which is still a matter of debate (see next section).

Whatever the procedure adopted, the lunar absolute model chronology is attained through a final calibration with the known radiometric ages. Unfortunately these are still largely incomplete and their attribution to specific geologic events is often uncertain (Stöffler and Ryder 2001; Stöffler et al. 2006). In particular, a large time interval from 3 to 1 Ga completely lacks data, several impact melts ages related to the formation of major basins are possibly contaminated by the single Imbrium event (Haskin et al. 1998), and the young ages related to the Copernican period are weakly defined (Stöffler and Ryder 2001) (Fig. 9).

Fig. 9
figure 9

MPF and NPF chronologies (from Marchi et al. 2009). The shadowed regions encompass a factor of ±2 around the MPF chronology curve

Full size image

Another important issue is represented by the secondary craters. Any production function assumes that the crater population considered is made up of randomly formed craters. However not all the cratering records meet this condition, since secondary craters are produced by fallback of high-velocity ejecta blocks from larger primary impacts and are strongly clustered in space and time (Melosh 1989; Wilhelms et al. 1987). The secondary craters can be often excluded from the statistics since they mostly occur in clusters, loops, lines, or display irregular rims, but small distant secondaries are difficult to detect and they may affect CSFDs for diameters less than 1 km (Shoemaker 1965; Wilhelms et al. 1978; Namiki and Honda 2003; Bart and Melosh 2007; McEwen and Bierhaus 2006). Actually, several lunar CSFDs are characterized by a steep branch at small crater sizes whose interpretation is still highly controversial, since some authors favour an important contribution of secondary craters (e.g. Guinness and Arvidson 1977; Wilhelms et al. 1978; Bottke et al. 2005; Dundas and McEwen 2007), whereas others consider their role almost negligible (Ivanov et al. 2002; Moore et al. 1980; Neukum et al. 1975; Neukum and Ivanov 1994; Ivanov 2006). However, the effect of secondaries is likely to be different as a function of the surface unit age, bedrock rheology and regolith cover; consequently, a detailed investigation of small craters on high resolution images represents a high priority objective.

7.2 Impact Flux

Any plot of crater density against absolute ages, derived from Apollo and Luna missions, highlights the fact that much of the extent of cratering on the Moon was produced during an early stage of its crustal evolution (Fig. 9; e.g. Wilhelms et al. 1987; Neukum et al. 2001; Marchi et al. 2009). In addition, on the basis of the absolute ages of impact melts, most of the major basins are thought to have formed in the time interval between 4.0 and 3.8 Ga (Cohen et al. 2000; Stöffler and Ryder 2001). All the evidence supports an event of Late Heavy Bombardment (LHB) characterized by a rapidly decaying flux. A heated debate concerns the LHB, which has been interpreted either as the declining tail of the primordial planetary accretion period or as a Lunar cataclysm that occurred at about 3.9 Ga (e.g. Neukum and Ivanov 1994; Hartmann et al. 2007; Chapman et al. 2007). These two interpretations imply contrasting scenarios: in the first one the LHB projectiles are assumed to be the long-surviving “left-over” planetesimals of the terrestrial planets formation zone (Hartmann et al. 1981; Neukum 1983), in the second asteroids were ejected from a primordial disk of small bodies at 16–36 AU by a presumed resonance sweeping due to the migration of Jupiter and Saturn (“Nice model” of Morbidelli et al. 2005; Gomes et al. 2005; Tsiganis et al. 2005). These two interpretations have severely different implications for Solar System formation as well as for the environment which facilitated the beginning of life on Earth. The major issues about the early impact flux and the time interval under which major impact basins were developed need more absolute age determinations on samples of certain attribution. In particular, ages are needed for Nectarian and pre-Nectarian basins with particular regard to the South Pole-Aitken basin which is believed to be the oldest large basin on the lunar surface (Wilhelms et al. 1987; Head et al. 1993) and which, also for this reason, will be one of the major observational targets of the MAGIA mission.

On the basis of the CSFD on the highlands and plains of the terrestrial planets, Strom et al. (2005) suggested that the LHB implies a MBAs source replaced in more recent times by an NEOs population source. This seems to be elegantly supported by the fit of the CSFD of several units of the Moon and Mercury surfaces with the recently proposed MPF (Fig. 10a) (Marchi et al. 2009; Massironi et al. 2009). Although all the evidence seems quite strong, the population of impactors through inner Solar System evolution is still unconsolidated, as demonstrated by the uncertain fit of pre-Nectarian terrain CSFD with any crater production function (Fig. 10b) (e.g. Neukum et al. 2001; Marchi et al. 2009) and the difficulties in quantifying the deviations in the recent NEO population caused by stochastic break-up occurrences which may have caused mass extinction events on the Earth (Bottke et al. 2000; Nesvorný et al. 2007, 2009). To solve these problems, as for those noted above, a much wider data set of absolute ages on samples of certain origin is required, as well as much more reliable observations of CSFD on Lunar surface units.

Fig. 10
figure 10

MPF and NPF best fits of CFSD: a Nectaris basin b Lunar highlands (from Marchi et al. 2009)

Full size image

7.3 The MAGIA Contribution to Lunar Chronology and Impact Fluxes

Most of the problems related to impact flux and crater chronology are due to an incomplete record of radiometric ages on samples of often uncertain stratigraphic attribution. Hence, a sustained campaign of sample return missions is needed and the related landing sites should be carefully selected to ensure reliable sample collection. To avoid any uncertainties in the attribution of future radiometric ages to specific geological events, detailed geological and stratigraphic mapping is needed. Regional and local coverage, respectively at moderate and high resolution, has proven to be necessary in order to define the distribution of geological units and their stratigraphic relationships (e.g. Wilhelms 1990). High resolution images show details of the regolith surface, stratigraphic contacts among bedrocks and cross-cutting relationships between structures, while images at moderate resolution show regional configuration of bedrocks, and the broad stratigraphic correlations and extension of the structures. In addition, the recent multi-spectral data from the Moon (Clementine mission, Kaguya-SELENE) demand a throughout integration and revision of the lunar geological maps in a way that details about composition or, at least, spectral characteristics of the surface are taken into account. In the polar regions MAGIA will be able to carry out a thorough geological investigation in full spatial and spectral resolution, satisfying on its own essentially all these geological mapping requirements. Moreover, it will provide a useful contribution also to the geological survey of the lunar farside by means of the planned CAM-SIR acquisitions on those regions.

The particular emphasis of MAGIA on acquiring data from the polar areas will provide also the basis for further crater counts on key units, which are likely to be representative of the early lunar history such as highlands and ejecta debris from ancient basins. In addition, new crater model ages for the limited volcanic deposits of the polar and far side regions may further constrain the highly heterogeneous lunar magmatism. The CARISMA high resolution images will also help specific studies on the production of secondary craters with the aim of assessing their contribution to the CSFD of the geological units at Lunar poles. Finally it may also contribute to analysing the effects of target physical properties on the CSFD of small craters.

8 Conclusion

Several recent missions to the Moon and numerous upcoming missions represent the first steps toward a new era of lunar exploration, most likely leading to sample return missions, rover exploration and human landings. The main objectives of the new era will mostly concern lunar geology and can be grouped into five closely interconnected fields: crustal composition; water; volcanism; tectonics; cratering record.

In each of these fields there are several questions still to be answered, in particular:

Crustal and Mantle Composition

What is the heterogeneous structure and composition of the lunar crust and mantle? What are the mechanisms of crustal and mantle evolution that lead to such a heterogeneous composition and asymmetrical structure?

Ice at the Lunar Surface

How did the volatiles on the lunar surface originate? What is their abundance and spatial distribution? What are the environmental, topographic and geological conditions favoring ice storage?

Volcanism

How did volcanism evolve during lunar geological history? How long did it last? What are the relations between volcanism and tectonics? How did the effusion rates change over lunar geological history? What are the implications for the thermal evolution of the Moon?

Tectonics

What are the relationships between basin evolution, lithospheric thickness and surface structures? What are the chronological and structural relationships between contractional cooling and basin related tectonics? How large is the radius contraction of the Moon?

Cratering Record

What was the nature, origin and duration of the LHB? How did the impact flux evolve through time? How can the existing crater chronology models be improved considering the various physical properties of the targets and an evolving impact flux? What was the contribution of secondaries to the CSFD and how much can they bias crater counts at different size ranges?

These major pending questions in lunar geoscience can be addressed by a complete data set including high resolution images from context cameras, synthetic views from medium resolution cameras, multispectral imaging, x- and gamma-ray mapping, topographic measurements from laser and radar altimeters, and 3D reconstructions from stereo acquisitions and radar sounding data. In this context, MAGIA can play an important role by providing data from a high resolution camera (CARISMA), an optical VIS camera and a VNIR imaging-spectrometer (CAM-SIR) and a Radar Altimeter (RASCAL). In particular, the unprecedented spectral resolution and range of CAM-SIR and the unique capabilities of RASCAL will yield data quality never obtained by previous missions. In addition, the lunar mission MAGIA will focus for the first time on an orchestrated effort to improve the understanding of the lunar polar regions.