Lunar Radar Sounder Observations of Subsurface Layers Under the Nearside Maria of the Moon

See allHide authors and affiliations

Science  13 Feb 2009:
Vol. 323, Issue 5916, pp. 909-912
DOI: 10.1126/science.1165988


Observations of the subsurface geology of the Moon help advance our understanding of lunar origin and evolution. Radar sounding from the Kaguya spacecraft has revealed subsurface layers at an apparent depth of several hundred meters in nearside maria. Comparison with the surface geology in the Serenitatis basin implies that the prominent echoes are probably from buried regolith layers accumulated during the depositional hiatus of mare basalts. The stratification indicates a tectonic quiescence between 3.55 and 2.84 billion years ago; mare ridges were formed subsequently. The basalts that accumulated during this quiet period have a total thickness of only a few hundred meters. These observations suggest that mascon loading did not produce the tectonics in Serenitatis after 3.55 billion years ago. Global cooling probably dominated the tectonics after 2.84 billion years ago.

Subsurface stratigraphic and tectonic features revealing the formation and evolution of a planetary body can be scanned by a ground-penetrating radar or radar sounder onboard a spacecraft. This was first realized by the Apollo Lunar Sounder Experiment (ALSE) onboard the Apollo 17 Command/Service Module (1), which revealed subhorizontal layering in Mare Serenitatis (2, 3). Recently, most of Mars has been probed by radar soundings (47).

The Lunar Radar Sounder (LRS) onboard the Kaguya spacecraft (SELENE) has been exploring the lunar subsurface since 20 November 2007. The LRS uses the HF band (5 MHz), enabling subsurface data to be obtained to a depth of several kilometers (8, 9). The LRS data reveal that most nearside maria have subsurface stratifications (Fig. 1). Many of the reflectors lie at apparent depths of several hundred meters below the surface. The deepest reflector was found in the northeastern Mare Imbrium at an apparent depth of 1050 m. Multiple-layer structures were observed in several mare regions, including the northeastern Imbrium, Crisium, and Oceanus Procellarum.

Fig. 1.

B-scan (10) displays with obvious subsurface echoes (red arrows). Doppler focusing was not applied to the data. Observation ground tracks are plotted on the topographic map (center left panel) that was generated using the LRS as an altimeter. A spherical surface with a radius of 1737.4 km is used as the reference for the vertical coordinates.

Subsurface echoes appear as apparent linear echo patterns a few hundred meters below the surface in B-scan display (10) (Fig. 2A and fig. S1). Similar echo patterns can also be produced by the range sidelobes of surface echoes. But the intensity of range sidelobes decreases so rapidly as their order increases that they should be negligible at a range of a few hundred meters from the main lobe. Spatially dense observations support the hypothesis that these echoes originate from subsurface horizons. Figure S2 shows B-scan displays of 40 adjacent orbit observations between 20° and 25.6°E. The constant appearance of these subsurface echoes indicates that there are subsurface boundary interfaces that reflect the LRS pulses. Estimating the dielectric constant of the lower subsurface-layer material provides a means for evaluating whether the detected echo is a subsurface reflection. Using a simple Fresnel reflection/refraction model (11), we estimated the loss tangent of the surface-layer material and the dielectric constant of the second-layer material (and the dielectric constant of lower-layer material if a second subsurface echo was detected) (Table 1) and found that they were consistent with the Apollo measurement results (12).

Fig. 2.

(A) B-scan display (10) down to an apparent depth of 2.5 km in Mare Serenitatis. (B) Same as (A) but down to an apparent depth of 7 km, indicating the absence of prominent reflectors at apparent depths of ∼2.7 and 4.7 km. Doppler focusing was not applied to the data.

Table 1.

Mare subsurface stratification at five locations, as inferred from LRS data. Apparent depth was determined by the mean of a number of A-scope data (10). The electric properties were estimated by inversely solving a Fresnel model with the assumption that the top-layer material has a dielectric constant of 4. The loss tangent, tan δ, of the lowermost layer was not determined, as the attenuation in that layer is unknown. d and D are the apparent depth and estimated depth to the upper surface of each layer, respectively; P is echo power.

View this table:

We attempted to identify the subsurface reflectors reported from the ALSE data for Mare Serenitatis (2). Two reflectors were inferred along east-west–trending tracks between 14° and 27°E at depths of ∼0.9 and 1.6 km when ϵ, the relative dielectric constant of lunar rocks, was assumed to be 8.7 (apparent depths of ∼2.7 and 4.7 km). Peeples et al. (2) distinguished subsurface echoes from noise by the multiple-orbit correlation technique, which compares echoes obtained from nearby orbits. However, no such reflectors were detected at apparent depths of ∼2.7 and 4.7 km in the LRS radargrams (Fig. 2B and fig. S1) for Serenitatis, although a more detailed analysis is necessary to conclusively show whether the ALSE reflectors exist. Instead, we found prominent reflectors lying at apparent depths of a few hundred meters (e.g., Fig. 2A and fig. S1). Because the ALSE had a range resolution of ∼1200 m in free space, it would be difficult to resolve shallow reflectors from the ALSE data. The actual resolution would be ∼400 m for bedrock with ϵ = 8.7; a lower value of ϵ would reduce this resolution. By contrast, the LRS has a range resolution of ∼75 m in free space, permitting detection of shallow reflectors.

Figure 3 and fig. S3 show profiles obtained by the LRS for Serenitatis, where the echo amplitudes were stretched to visualize subsurface features. Subhorizontal prominent reflectors are seen at apparent depths of several hundred meters. Linear topographic features, such as mare ridges parallel to the nadir tracks, can generate linear echo traces similar to those of subhorizontal reflectors in B-scan images, but we eliminated this possibility by estimating the off-nadir distances of the features in question.

Fig. 3.

LRS data along a north-south–trending track in southern Serenitatis. (A) Red and green lines show the nadir track of Kaguya and the boundaries of stratigraphic units S11, S15, S22, and S28 defined by (14). Solid triangles indicate the boundary between S11 and the other units. S15 and S22 are bounded largely by Dorsum Nicol and are separated from S28 by Dorsa Lister. [Apollo image AS17-M-0452] (B) LRS data along the track. This shows a close-up of the southern part of profile c-c′ in fig. S3. (C) Prominent reflector found in the profile. This reflector appears at the lunar surface at a latitude of ∼17.8°. The apparent dip of the reflector is ∼3.1° in the vicinity of this outcrop, giving a true dip of 1° to 2° when the dielectric constant of the rocks is accounted for. This value is consistent with the surface slope to the south of this latitude.

The subsurface reflectors are predominantly horizontal, which suggests that they are interfaces within the mare fill; the base of the fill should have a topographic relief as large as that of the surface of Orientale, an impact basin approximately the same size as Serenitatis (13). Mare volcanism peaked at 3.6 to 3.7 billion years ago (Ga), several hundred million years after the basin-forming impact event (14). The heavy bombardment during this period should have enhanced the undulations. Thus, the reflectors are considered to indicate not the base of mare deposits, but layers within mare deposits. Thus, it is important to consider the stratigraphy of mare units evaluated by Hiesinger et al. (14). They subdivided the mare into spectrally homogeneous lithologic units, the ages of which were determined by crater counting.

The lithology responsible for these reflectors is unclear, but high-porosity deposits, such as buried regolith and pyroclastic and ejecta deposits, are possibilities. However, pyroclastic and ejecta deposits can be ruled out on the basis of the following observations. First, the reflectors have horizontal dimensions of >200 km. Second, layers represented by some reflectors appear to outcrop at the lunar surface. Also, the strata above and below the reflectors can be correlated to geological units on the surface. An excellent example is the wavy reflector in Fig. 3, which dives at ∼17.8°N and extends to 24° to 25°N under the surface. The intersection between this reflector and the surface coincides with the boundary between mare unit S11 (14) and its overlying strata. Except at the margins of Serenitatis, the apparent depth of this horizon is 400 to 500 m (Fig. 3B), which corresponds to an actual depth of ∼250 m or shallower if the dielectric constant is assumed to be 4 or higher (12). The reflector of Serenitatis in Table 1 is correlated by its apparent depth with the prominent reflector in profile d-d′ (fig. S3), which is adjacent to the locations given in Table 1. Thus, the top of S11 lies at an actual depth of ∼175 m at around 22.0°N, 23.3°E. Hiesinger et al. (14) used crater chronology to estimate that this unit formed at 3.55 Ga, whereas the overlying units S15 and S22 were dated at 3.44 and 3.28 Ga, respectively. Considering that the apparent depth is overestimated by a factor of ∼2, we see that the reflector identified with the top of S11 to the south of 18°N is an extension of the slope made by the surface of this unit to the south of ∼17.8°N, where the reflector appears at the surface (Fig. 3C).

The depositional hiatuses represented by the reflectors may have lasted 0.17 to 0.11 billion years, which would allow the regolith to be thicker than 1 m (15). Accordingly, we interpret this reflector as being the buried regolith blanketing S11, although the thickness of the layer returning the prominent echoes remains unknown. An identical interface appears at a similar apparent depth in profile e-e′. This interface meets the surface at ∼19°N, where S11 is overlain by S15 (fig. S3). In addition, S28 lies above the reflector. Consequently, the layers above S11 have ages between 3.56 and 2.84 Ga. Pyroclastic deposits are unlikely to be the source of the prominent reflectors that meet the surface, because there is no regional pyroclastic deposit with an age between 3.55 and 2.84 Ga in or around Serenitatis (16).

Ejecta blankets are unlikely to account for the extensive reflectors. Continuous ejecta extend about one crater radius from the crater rim (17). Thus, an ejecta blanket that is the same width as the reflectors would be accompanied by a basin-sized crater, but the tracks extend beyond the continuous ejecta from the large craters produced between 3.55 and 2.84 Ga. The top of S11 is at an actual depth of <250 m. Thus, the relationship between crater size and rim height (18) indicates that basin-sized craters cannot be concealed by strata above this depth.

The reflector identified with the top of S11 extends from the mare's southern margin to its central region. The deeper prominent reflector in profile a-a′ (fig. S3) is correlated with the top of this unit, because the apparent depth of this reflector is generally consistent with that of the interfaces identified with the horizon found in the LRS data from neighboring orbits. The basalts of S11 are known to have a higher TiO2 content and a lower reflectivity than younger ones (Fig. 3) (19, 20). The ejecta from the Bessel crater (21.8°N, 17.9°E) to the west of Deseilligny have similar spectral characteristics to those of S11, which suggests that the unit lies under the central region of Serenitatis (21).

The horizon at the top of S11 is equivalent to the interface between units I and III in the nomenclature of Solomon and Head (22), who estimated that the thickness of unit III was 1 km at the margin of the basin and 2 km at its center. They estimated that the total mare fill was 8.5 km thick at the center by assuming that before mare flooding, the basin had a topography similar to that of Orientale. They also estimated the thickness of the mare fill from the gravity anomaly. In contrast, the thicknesses of basalt fills in many maria, as inferred from partially flooded craters, are less than estimates obtained by assuming a prefill topography similar to the present topography of empty basins. De Hon et al. (23) showed that nearside mare fills are generally thinner than 1 km. The thinner estimate was recently verified by Clementine spectral data from the Humorum basin (24). Although the thickness of the mare fill under the top of S11 is unknown, the thickness above this horizon determined by the LRS is inconsistent with the thicker estimate but is closer to the thinner one.

The ALSE data were used to support the tectonic origin of mare ridges [e.g., (25)]. Peeples et al. (2) illustrated an anticline marked by reflectors under a mare ridge in Serenitatis. Although the existence of these reflectors was not verified by the LRS, our data support the hypothesis. The wavy reflector in Fig. 3 and that between 18° and 21°N in profile d-d′ (fig. S3) are subparallel to the surface undulations, which are parts of ridge systems in southern Serenitatis (25). The northward continuation of the latter reflector is disturbed by Dorsa Smirnov, a broad linear topographic swell. Thus, basaltic layers are folded under mare ridges.

The thicknesses of the strata above S11 appear to be independent of their positions relative to the ridges; this suggests a tectonic quiescence during the emplacement of the younger units (i.e., from 3.55 to 2.84 Ga). Mare basalts had very low viscosities (26), resulting in approximately horizontal lava fields. Their original horizontality should have resulted in thinning of strata toward anticlines if the folds grew during the deposition of the strata (fig. S4). Therefore, post-S11 strata were subject to horizontal shortening to form mare ridges predominantly after the emplacement of S28 (2.84 Ga).

The tectonic quiescence and the thin post-S11 strata are in conflict with the tectonic model of mascon loading (22). This model predicts that the accumulation of strata gradually bends the lithosphere by their weight, giving rise to syndepositional tectonics. Accordingly, mare fills were assumed to be several kilometers thick. The superstrata over S11 were thought to be 1 to 2 km thick to have sufficient weight for the model (22). However, Freed et al. have demonstrated the difficulty of explaining mare ridge formation after S11 emplacement by the strata, even if they were assumed to be 2 km thick (27). Instead, they conjecture that the mare ridges were formed by a combination of loading and compression due to global cooling. Our study reveals that the superstrata are about one-tenth the thickness of those predicted by the model. Thus, mascon loading was not responsible for forming the ridges after 3.55 Ga, and global cooling was probably the dominant cause of post–3.55 Ga tectonics, at least in southern Serenitatis.

Supporting Online Material

Figs. S1 to S4

References and Notes

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article