Gravity field of the Orientale basin from the Gravity Recovery and Interior Laboratory Mission

See allHide authors and affiliations

Science  28 Oct 2016:
Vol. 354, Issue 6311, pp. 438-441
DOI: 10.1126/science.aag0519

On the origin of Orientale basin

Orientale basin is a major impact crater on the Moon, which is hard to see from Earth because it is right on the western edge of the lunar nearside. Relatively undisturbed by later events, Orientale serves as a prototype for understanding large impact craters throughout the solar system. Zuber et al. used the Gravity Recovery and Interior Laboratory (GRAIL) mission to map the gravitational field around the crater in great detail by flying the twin spacecraft as little as 2 km above the surface. Johnson et al. performed a sophisticated computer simulation of the impact and its subsequent evolution, designed to match the data from GRAIL. Together, these studies reveal how major impacts affect the lunar surface and will aid our understanding of other impacts on rocky planets and moons.

Science, this issue pp. 438 and 441


The Orientale basin is the youngest and best-preserved major impact structure on the Moon. We used the Gravity Recovery and Interior Laboratory (GRAIL) spacecraft to investigate the gravitational field of Orientale at 3- to 5-kilometer (km) horizontal resolution. A volume of at least (3.4 ± 0.2) × 106 km3 of crustal material was removed and redistributed during basin formation. There is no preserved evidence of the transient crater that would reveal the basin’s maximum volume, but its diameter may now be inferred to be between 320 and 460 km. The gravity field resolves distinctive structures of Orientale’s three rings and suggests the presence of faults associated with the outer two that penetrate to the mantle. The crustal structure of Orientale provides constraints on the formation of multiring basins.

Basin-scale impacts disrupted early planetary crusts and imparted substantial kinetic energy that led to fracturing, melting, vaporization, seismic shaking, ejection, and redistribution of crustal and possibly mantle material. Impact basins preserve the record of these planet-altering events, and the study of these structures elucidates the partitioning of energy and its corresponding geological and environmental effects early in planetary history. Surface signatures of impact basins on solid planets have been extensively documented (1, 2), but their subsurface structure has, to date, been poorly characterized. We present a high-resolution orbital gravity field model of the Orientale basin on the Moon as mapped by the Gravity Recovery and Interior Laboratory (GRAIL) mission (3).

Orientale, located on the southwestern limb of the lunar nearside, is the youngest (~3.8 billion years old) (1, 4) large (~930-km diameter) impact basin on the Moon. As a consequence of its good state of preservation (1, 5), with relatively few superposed large craters (6), it is often considered the standard example of a well-preserved, multiring basin in comparative studies of large impacts on terrestrial planetary bodies (2, 7). Because of the basin’s importance, the GRAIL Extended Mission (see the supplementary text) featured a low-altitude mapping campaign during the mission’s Endgame phase (8), in which the dual spacecraft orbited the Moon at an average altitude of 6 km and acquired observations less than 2 km above the basin’s eastern rings (figs. S1 and S2).

To produce the highest-resolution gravity map achievable from the data and to ensure that small-scale features resolved were robust, we developed two maps that used the same data but independent methodologies (9). The first is derived from a global spherical harmonic expansion of GRAIL’s Ka-band (32 GHz) range-rate (KBRR) tracking data to degree and order 1200 (spatial block size = 4.5 km). The second is from a local model that implemented a short-arc analysis (10) of the tracking data and used a gravitational field model to degree and order 900 (11) as the a priori field. Local gravitational anomalies were estimated with respect to the spherical harmonic model at the center coordinates of a set of grid points. The final model has a mixed-grid resolution of 0.1° by 0.1° and 0.1667° by 0.1667°, corresponding to a maximum spatial resolution varying between 3 and 5 km. These independent analyses produced gravitational models of Orientale that are essentially indistinguishable (see fig. S4). The maps are shown in Fig. 1; they resolve the shallow subsurface structure of Orientale at a spatial resolution comparable to that of many geological structures at the surface, including simple and secondary craters.

Fig. 1 High-resolution maps.

(A) Topography, (B) free-air anomaly (1 mGal = 1 milliGalileo = 10−5 m s−2), (C) Bouguer anomaly, (D) crustal thickness over shaded-relief topography, and (E) Bouguer gravity gradient (1 Eötvös = 10−4 mGal m−1 = 10−9 s−2) of the Orientale basin and surroundings. Dashed lines in (A) from innermost to outermost correspond to the Inner Depression, Inner Rook Ring, Outer Rook Ring, and Cordillera Ring (ID, IRR, ORR, and CR, respectively). The solid white line in (D) shows the location of the cross-sectional profile A–A′ in Fig. 2A. Blue lines show the locations of the azimuthally averaged cross sections in Fig. 2B. Topography is updated from Lunar Orbiter Laser Altimeter (LOLA) (12) map LDEM_64, 0.015625° spatial resolution. To highlight short-wavelength structure, we have subtracted spherical harmonic degrees and orders less than 6 from the Bouguer gravity field. Calculation of crustal thickness and Bouguer gravity gradient is discussed in the supplementary materials (9).

The topography of the Orientale basin (12) from the Lunar Orbiter Laser Altimeter (13) and the free-air gravity anomaly field of the region are shown in Fig. 1, A and B. The maps show similar detail at small spatial scales because above degree and order 80 (spatial block size <68 km), more than 98% of the lunar gravity field is attributable to topography (14). The high correlation of topography and gravity at short horizontal scales is due to the large magnitude of the gravity anomalies arising from topography relative to the weaker anomalies attributable to density anomalies in the shallow subsurface (14).

Both topography and free-air gravity anomaly resolve Orientale’s Inner Depression, as well as the Inner Rook ring, Outer Rook ring, and Cordillera ring (see Fig. 1). The rings, which were only partially resolved in pre-GRAIL gravitational models (15), formed in the process of cavity collapse during the modification stage of the impact event, within an hour of the initiation of basin formation (16). The mechanism for ring formation, however, has been controversial (2, 5, 1719), in large part because of a lack of understanding of subsurface structure needed to provide constraints on impact basin formation models.

Variations in subsurface mass are best revealed in the Bouguer gravity anomaly field (Fig. 1C), a representation of the gravitational field after the attraction of surface topography has been removed. Determination of crustal structure requires careful consideration of likely crustal and mantle composition (supplementary text). For uniform-density crust and mantle [2550 and 3220 kg m−3, respectively (20)], the Bouguer gravity anomaly can be used to map the crust-mantle boundary and thus crustal thickness (Fig. 1D and fig. S5). Although the assumption of uniform density is an approximation, its application to the regional crustal structure is supported by crustal density inferred from GRAIL (20), as well as from orbital remote sensing data (supplementary text and fig. S6). Some models for the crust invoke a mixed feldspathic layer that overlies a layer of pure anorthosite (21), but the density contrast between these rock types is small in comparison with that across the crust-mantle interface. In the mantle, there is likely a pronounced seismic discontinuity at 500- to 600-km depth, about the scale of Orientale, marking a change in mantle composition and hence density (22); lateral variations in this depth will contribute to the Bouguer signal. However, the density contrast is provided by a change in mafic mineralogy and is small compared with the density change at the crust-mantle boundary. Furthermore, the signal is attenuated to the extent that only the broadest length scales of the crustal model might be affected.

A cross section of crustal structure along the profile in Fig. 1D is shown in Fig. 2. Crust-mantle boundary profiles take into account the effect on the crustal structure of a 10-km-thick sheet of density 2650 kg m−3 confined to the inner depression to represent solidified impact melt or crustal material formed by differentiation of Orientale’s melt pool (9); in the end-member case, in which the sheet is instead identical in density to feldspathic crust, the minimum crustal thickness would be only 2 km less. Consequently, the presence of a high-density sheet does not have a substantial effect on the basin’s crustal structure. Figure 1D indicates that the Orientale impact removed and redistributed a minimum of (3.4 ± 0.2) × 106 km3 of material (9) from the lunar crust. Approximately one-third of the excavated material was deposited as ejecta in an annulus between the Cordillera ring and a radial distance from the basin center of one basin diameter (23), contributing to enhanced crustal thickness in this region (Fig. 1D). The high porosity (~18%) inferred in the Orientale ejecta blanket (20, 24) is similar to the porosity observed in Apollo 14 samples of the Fra Mauro Formation, which consist of ejecta from the Imbrium basin (25).

Fig. 2 Crustal cross section.

(A) Subsurface structure of the Orientale basin along the profile shown in Fig. 1D, from southeast to northwest. Crust is shown as tan, melt sheet as red, and mantle as green. Arrows above the cross section denote, inward to outward, Orientale’s ID, IRR, ORR, and CR. The heavy solid line indicates the base of the crust in the presence of a melt sheet that is 10 km thick, 350 km in diameter, and 2650 kg m−3 in density; the thin solid line indicates the base of the crust if the melt-sheet density is identical to that of the crust. Vertical exaggeration (V.E.) is 5:1. (B) As in (A), with no vertical exaggeration and filter for downward continuation (9) having higher resolution by a factor of 3; the profile is azimuthally averaged in sectors (azimuth measured clockwise from north; see Fig. 1D) to suppress noise. Red lines give the locations of proposed faults dipping inward at 50° dip angle from the nominal surface positions of the ORR and CR. Other variations in crust-mantle boundary depth suggest the presence of additional faults.

The calculation of crustal thickness does not include the presence of the mare units that are generally less than 1 km thick and irregularly distributed within the Inner Depression and ponded irregularly inside the Outer Rook and Cordillera rings (21) (low-reflectance areas in fig. S7). The gravitational signal of the mare units is partially masked by the unknown characteristics of the impact melt sheet, and where these deposits are present, the underlying crustal thickness would be slightly greater than in the model.

Pre-GRAIL gravitational models have shown that the most distinct, well-preserved, large-scale gravitational signature associated with large lunar basins is a strong central positive free-air anomaly (2629). These mass excesses, or mascons (30), are a consequence of the thinning of the crust beneath the basin combined with mantle rebound and subsequent isostatic uplift of the central basin that was mechanically coupled to surrounding crust (31, 32). Figures 1 and 2 show that in Orientale, the diameter of crustal thinning corresponds closely to the diameter of the Inner Depression. We therefore identify the Inner Depression as an approximation to the basin excavation cavity—i.e., the region in which the crust was thinned from its pre-impact state (21).

Of interest in impact studies is the transient crater—the cavity formed at the culmination of excavation, before collapse and modification. The transient crater corresponds to the maximum displaced volume, and its diameter satisfies standard impact scaling laws (33, 34), which allow many characteristics of an impact to be inferred from its size. Although there is general recognition that the transient crater should not be preserved, each of Orientale’s basin rings has been variously interpreted as marking the diameter of the transient crater (2, 5, 17, 18, 35). GRAIL’s improved spatial resolution and signal sensitivity show no evidence of the transient crater, indicating that any remnant is not preserved or at least not recognizable in the gravity field. However, the transient crater diameter may now be calculated with confidence from scaling laws (36). From the 530-km diameter of the annular maximum in crustal thickness (Fig. 1D), measured from an azimuthally averaged crustal thickness profile (fig. S5), the transient crater diameter is between 320 and 460 km, placing it between the diameters of Orientale’s Inner Depression and Inner Rook ring. The transient crater thus does not correspond to a specific ring; indeed, hydrocode modeling constrained by this crustal structure model (16) indicates that rings form subsequent to the transient crater, during the collapse phase.

The transition between the basin excavation cavity and the surrounding crust is well illustrated in Fig. 2. At the outer edges of the zone of mantle uplift, the crust-mantle boundary slopes outward and downward by at least 20° to 25°. The spatial correspondence of this plug of uplifted mantle with the Inner Depression is similar to the pattern seen in other multiring basins (37), but it is in contrast to peak-ring basins, where the zone of uplifted mantle is limited to the area within the peak ring.

The model also shows, beyond the basin depression, an annulus of thickened crust (Fig. 1D and fig. S8D), as well as radial structure in gravity gradients (Fig. 1E and fig. S8E) that locally correlates with observed ejecta structures (e.g., secondary crater chains) (23).

Aspects of Orientale’s asymmetry in surface structure extend to the subsurface, as indicated in Figs. 1 and 2. For instance, the basin exhibits an east-west variation in regional crustal structure that predated formation of the basin.

There are also radial variations in crustal thickness, some of which are distinctly associated with Orientale’s outer two basin rings. The simplest interpretation of the azimuthally averaged models is that they could correspond to displacements associated with normal faults that penetrate the crust. The crust-mantle boundary relief in Fig. 2B suggests that there could be multiple faults dipping inward from the Outer Rook and Cordillera rings. The crustal thickness model also suggests the presence of other crustal faults that lack a visible surface expression. Although these faults may be listric—i.e., the dip angle may decrease with depth—a dip of 50°, indicated by hydrocode simulations (16), is consistent with prominent changes in crust-mantle boundary depth. These simulations also support crustal faulting not associated with rings.

Insight into the distinctive nature of each ring can be gained from scrutiny of Figs. 1 and 2 and fig. S8. The Inner Depression has the most axisymmetric and the largest variation in crustal thickness; a change in the sign of the gravity gradient (9) marks the depression’s edge. The topography of the Inner Rook ring is morphologically similar to that of peak rings in small basins (16). Individual peaks within the Inner Rook are associated with positive free-air and Bouguer anomalies embedded within an annulus of negative free-air and Bouguer anomalies. The Inner Rook also appears associated with a near–circumferentially continuous change in the sign of the gravity gradient (Fig. 1E) and a flattening in relief along the crust-mantle boundary.

The Outer Rook ring also displays well-developed topography consistent with the surface expression of a normal fault scarp (38). Ring topography has associated positive free-air anomalies embedded within the same annulus of negative free-air and Bouguer anomalies. The most negative Bouguer gravity in the region appears within the Outer Rook and may reflect a combination of thickening of the crust by ejecta and extensive fracturing in the crustal column. The Outer Rook displays a sign change in the gravity gradient and a mild shoaling of the crust-mantle boundary. The collective characteristics of the Outer Rook ring are consistent with local thinning of the crust by faulting.

The topography of the Cordillera ring deviates markedly from axisymmetry; it is less developed than the Inner and Outer Rook rings and has little expression in part of the basin’s southwestern quadrant. This asymmetric structure may be a consequence of the northeast-to-southwest–directed oblique impact that formed the basin (39) or preexisting heterogeneity of crustal or lithospheric structure (2, 38), with a clear west-to-east gradient of decreasing crustal thickness still preserved (Fig. 1D). The topography of this ring is also consistent with the surface expression of a normal fault scarp (16). The ring is characterized by positive free-air and Bouguer anomalies, a gradient in crustal thickness, and a circumferentially discontinuous sign change in the gravity gradient. The variation of relief along the crust-mantle boundary strongly suggests fault penetration to the lower crust and possibly upper mantle. The gravitational signature could alternatively reflect contributions from magmatic intrusions along the irregularly developed ring fault, but regional seismic reflection profiles of a portion of the terrestrial Chicxulub impact structure, 20 to 25% the size of Orientale, show ring faults that extend well into the lower crust (40).

Our observations, combined with the accompanying simulations (16), elucidate the planet-shaping thermal, tectonic, and geological consequences of Orientale and, by extension, other large impacts that dominated the early evolution of the Moon and other solid planets.

Supplementary Materials

Materials and Methods

Supplementary Text

Figs. S1 to S8

References (4175)

References and Notes

  1. Materials and methods are available as supplementary materials on Science Online.
Acknowledgments: The GRAIL mission is supported by NASA’s Discovery Program and is performed under contract to the Massachusetts Institute of Technology and the Jet Propulsion Laboratory. Topography was obtained from the Lunar Orbiter Laser Altimeter on the Lunar Reconnaissance Mission, managed by NASA’s Goddard Space Flight Center. The NASA Pleiades and Center for Climate Simulation supercomputers were used to compute the gravity solutions. All data used in this study are archived in the Geosciences Node of the NASA Planetary Data System at

Stay Connected to Science

Navigate This Article