Geology of the Smythii and Marginis region of the Moon' Using integrated remotely sensed data

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 105, NO. E2, PAGES 4217-4233, FEBRUARY 25, 2000 Geology of the Smythii and Marginis region of the Moon' Using integrated remotely sensed data Jeffrey J. Gillis Department of Earth and Planetary Sciences, Washington University, Saint Louis, Missouri Paul D. Spudis Lunar and Planetary Institute, Houston, Texas Abstract. We characterized the diverse and complex geology of the eastern limb region of the Moon using a trio of remote-sensing data sets: Clementine, Lunar Prospector, and Apollo. On the basis of Clementine-derived iron and titanium maps we classify the highlands into low-iron (3-6 wt % FeO) and high-iron (6-9 wt % FeO) units. The association of the latter with basalt deposits west of Smythii basin suggests that the highland chemical variation is the result of mixing between basalt and highland lithologies. Mare Smythii and Mare Marginis soils are compositionally similar, containing moderate iron (15-18 wt % FeO) and titanium (2.5-3.5 wt % TiO2). Smythii basin, in addition to the basalt deposits, contains an older, moderate-albedo plains unit. Our investigation reveals that the dark basin plains unit has a distinct albedo, chemistry, and surface texture and formed as a result of impact-mixing between highland and mare lithologies in approximately equal proportions. Clementine iron and maturity maps show that swirls along the northern margin of Mare Marginis have the same iron composition as the surrounding nonswirl material and indicate that the swirl material is bright because of its low agglutinate content. Gravity data for the eastern limb show high, positive Bouguer gravity anomalies for areas of thin basalt cover (e.g., Smythii basin and complex craters Joliot, Lomonosov, and Neper). We deduce that the uplift of dense mantle material is the primary (and mare basaltic fill the secondary) source for generating the concentration of mass beneath large craters and basins. 1. Introduction The Smythii-Marginis region is one of the most geologically diverse areas on the Moon (Figure 1) [Spudis and Hood, 1988]. We have studied the eastern limb region of the Moon from 70 ø to 100 ø longitude and from 10øS to 30øN utilizing Lunar Orbiter and Apollo photographs, Apollo y- and X-ray data [e.g., La Jolla Consortium, 1977], Clementine multispectral, gravity, and topography data [e.g., Nozette et al., 1994], and recently obtained Lunar Prospector gravity and y-ray data [Binder, 1998]. Three primary geologic units are found in this area: cratered highlands, characterized by high albedo and low FeO (-3-10 wt % FeO); basin plains material, characterized by moderate albedo and FeO content (6-11 wt %); and mare material, characterized by low albedo and high FeO content (15-18 wt %) (Figure 2; Plate 1; Table 1). The regional geology is controlled by pre-nectarian basins, from the oldest and most degraded basins (e.g., Balmer, Marginis, Lomonosov/Fleming, and Al-Khwarizmi/King) to the relatively well-preserved Smythii basin with its rugged inner rim. Feldspathic highlands material is the dominant unit exterior to the basins. The mare and dark basin material are contained within the Smythii and Marginis basins and large crater such as Joliot (25øN, 94øE; 170 km diameter), Neper (2øS, 85øE; 140 km diameter), Lomonosov (27øN, 94øE; 95 km diameter), and Hubble (22øN, 87øE; 85 km diameter) (Figures 1 and 2). Copyright 2000 by the American Geophysical Union. Paper number 1999JE001111. 0148-0227/00/1999JE001111 $09.00 4217 The geology of the eastern limb region contains valuable information about primary and secondary crustal formation processes on the Moon. Numerous multiring basins and large craters on the eastern limb have sampled highland terra material from different depths in the crust, the depth of excavation being proportional to the diameter of the crater [e.g., Spudis, 1993]. Using the composition of the ejecta blankets, central peaks, and basin rims, we have constructed the three-dimensional compositional structure of the highlands crust on the eastern limb of the Moon. Studying the duration and timing of magmatism in this region will aid in understanding the volcanic and thermal history of the Moon. Volcanism in the area has occurred in at least three dis- tinct episodes. Observations of the crater densities for the volcanic flows in Smythii suggesthat these flows are among the youngest on the Moon [Schultz and Spudis, 1983; Spudis and Hood, 1988]. The basalt deposits in Mare Marginis have a higher density of craters and are therefore older. Finally, dark halo craters on the floors of the Lomonosov/Fleming and Al- Khwarizmi/King basins are evidence of buried basalt deposits [Schultz and Spudis, 1979, 1983], indicating that volcanism occurred before the end of heavy bombardment,-3.85 Gyr ago. Mare basalts fill most of the nearside basins, thus erasing the early stages of basin-filling volcanism. Smythii basin, because it is only partly filled, affords the opportunity to study the early stages of volcanic basin filling. In addition to the mare unit in Smythii, the basin floor consists of a moderate albedo plains unit (Figures 1, 2, and 3). The origin of this unit is hypothesized as either a mixture of highlands and volcanic material [Conca and Hubbard, 1979] or ejecta from the Crisium basin [Stewart et al., 1975]. We will demonstrate that a fifty-fifty mixture of highlands and mare material correctly matches the composition and

4218 GILLIS AND SPUDIS' GEOLOGY OF THE MOON'S EASTERN LIMB Figure 1. This Apollo image shows a regional view of the eastern limb of the Moon (10øS-40øN latitude and 70ø-100 ø longitude). Extending across much of the region are the crater rays of Giordano Bruno (GB), one of the youngest craters on the Moon. Large craters partly filled with basalt are Lomonosov (L), Joliot (J), Hubble (H), and Neper (N). Lunar swirls (LS) north of Marginis basin are associated with large surface magnetic anomalies. Mare-filled basins are Mare Marginis (MM) and Mare Smythii (MS). Note how the dark basin plains (dbp) material is almost as dark as the basalt in Smythii and maintains an arcuate edge along the southwestern inner ring of Smythii (AS14-75-10314). albedo of the Smythii basin floor material. Furthermore, the mafic composition of the ejecta from craters on the basin floor unit, and its relationship with surrounding units, is evidence of a highland-mare mixture. 2. Clementine Data Processing Clementine multispectral images from 21 orbits obtained during the first and second months of data acquisition covering the eastern limb region were used in this study (Figure 2). Im- ages from the three band passes (415, 750, and 950 nm) of the ultraviolet-visible (UV-VIS) camera were processed using Integrated Software for Imaging Spectrometers (ISIS), developed by the U.S. Geological Survey (USGS), Flagstaff, Arizona [Gaddis, 1996; Eliason, 1997]. All filters were coregistered to the 750 nm filter to align surface features in each of the three bands and prevent pixel offset when using the images for quantitative calculations. Calibration parameters developed by the USGS, Flagstaff (e.g., gain, offset, dark-field subtraction, fiat-field, frame transfer, and exposure-time corrections), were used to radiometrically

GILLIS AND SPUDIS' GEOLOGY OF THE MOON'S EASTERN LIMB 4219 lo Figure 2. Clementine 750 nm image mosaic showing the eastern limb region under high Sun illumination. For part of orbit 268, data were not obtained at the 750 nm wavelength' thus 900 nm data from that orbit were forced to fit the reflectance of the 750 nm data in order to increase surface coverage. Images from this orbit are not included in any of the calculations. This image is in sinusoidal projection and covers between 10 south to 30 north and 75 to 95 east.

4220 GILLIS AND SPUDIS' GEOLOGY OF THE MOON'S EASTERN LIMB 80OE 90øE FeO Wt.øo >18 16 14 12 lo 8 6 4 2 10 ø e ø 70øE 80øE 90øE.10 ø Plate 1. FeO map of the Smythii-Marginis region. The iron content for the basalt material within the two basins is similar. The moderate albedo unit within Smythii basin is much lower in iron than the basalt unit.. The highland terra to the west of Smythii basin is higher in iron relative to the highlands east of the basin

GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB 4221 Table 1. Units Within the Eastern Limb Region Defined by Their Chemical Signature Unit FeO+l TiO2+l A1203 wt % MgO wt % Th (ppm) 'Fh(ppm) wt '70 wt % Apollo Apollo Apollo Prospector Highlands East 3-7 < 1 28-29 4-5 0.5 0.7 Highlands West 6-10 < 1 26-27 8.3-9.6 0.7 1.4 Mare Smythii 16-18 2.5-3.5 16-22 9.9-12.4 2.4 1.4 Basin plains 6-11 0.75-1.8 22-25 9.6-10.3 1.2-1.6 1.5 Mare Marginis 15-17 2.6-3.6 1.2 FeO and TiO2 concentrations are derived from the Clementinc 415,750 and 950 nm data [Lucey et al., 1998]. Apollo x-ray data yield A1203 and MgO abundances [Conca and Hubbard, 1979], and Th is from the Apollo )'-ray data [Davis, 1980] and Prospector ),-ray data [Lawrence et al., 1998; Gillis et al., 1999]. correct the data set (convert 8-bit digital numbers to 32-bit absolute reflectance (I/F)) [McEwen e! al., 1998]. The photometric function by McEwen [1996] and revisions by McEwen et al. were not obtained during a previous orbit, 111, as a result of a computer reset which prevented Clementinc from recording data. The data described above were not used in the construction of the [1998] were then applied to normalize images to a standard regional mosaic; their absence accounts for the "checkerboard" viewing geometry (phase=30 ø, emission=&, incidence=30 ø) to allow direct comparison of reflectance values. Final image mosaics were created with a resolution of 250 m per pixel in sinugaps evident in Figure 2 and Plates 1 and 2. Furthermore, we have not processed the 900 or 1000 nm data, as their inclusion into the mosaic increases the size and number of these data gaps. soidal map projection. The first month's data, from 5øS through 10øN, were obtained with phase angles of 5 ø or less. As a result, reflectance values for the first month's data are higher (_<4%) than the reflectance val- 3. Smythii Basin 3.1. Ring Structure ues for the overlapping second month's data after photometric Smythii basin (2øS, 87øE) is pre-nectarian [El-Baz and normalization. We assumed that the second month's data better Wilhelms, 1975]. There exists some debate concerning where the represent the true reflectance of the lunar surface because of their basin ring structures are located. Wilhelms [1987] mapped four greater phase angles (>20ø). This conclusion was made on the ring structures with diameters of 360, 660, 840, and 960 km. basis that photometric models are most accurate at moderate Spudis [1993], however, suggested five basin rings: 260, 320, phase angles and that the photometric normalization modifies 540, 740, and 1130 km in diameter. The topographic basin rim reflectance values less for an image with a phase angle of 20 ø diameter is reported as 740 km [Spudis, 1993] or 840 km than for an image with a phase angle of 5 ø An empirical fit was [Wilhelms, 1987]. On the basis of Clementinc altimetry (Figure performed on each filter in the first month's data between 5øS 4), and Lunar Orbiter and Apollo images, we find evidence for and 10øN in order to match the reflectance data in overlapping three rings with diameters of 360, 660, and 840 km. The toporegions from the second month's orbits. graphically highest of the three structures is the 840 km ring; We used a three-channel, false-color composite mosaic to estherefore we consider it the main topographic rim. timate qualitatively the regolith composition and glass content. Maps of estimated FeO (Plate 1) and TiO 2 were created using the method developed by Lucey e! al. [1995, 1996, 1998] and with refinements by Blewerr et al. [1997]. Excellent agreement between the spectral Fe and Ti parameters and the average FeO and TiO 2 contents from the Apollo and Luna sampling sites demonstrates that these techniques can confidently be applied to the Clementinc UVVIS coverage of the Moon globally [Blewerr et al., 1997]. Image gaps occur in the Clementinc mosaics for the eastern limb region for multiple reasons. In eight out of the 10 second month's orbits (264-266, 268, 269, 271,273), images from one of the three filters used to construct either the color ratio irnage or iron or titanium abundance maps (415, 750, or 950 nm) was 3.2. Highland Material The innermost ring of Smythii stands 4 km above the basin floor (Figure 4) and represents both a physical and chemical boundary between the inner and outer basin. Plagioclase-rich highland material is the dominant geologic unit outside the innermost ring of Smythii. This material is mostly ejecta from the Smythii basin, mixed with the ejecta of nearby younger basins and large craters. The low-iron (3-7 wt % FeO; Plate 1) and high-alumina content (Table 1; 26-29 wt % A1203 [Conca and Hubbard, 1979]) of the soils in this area suggests that the crust is thick and highly feldspathic in composition to the depth sampled by the Smythii basin impact (-30 km). However, the highlands dropped or the images obtained were >15 ø off nadir (orbits 264 west and north of the basin exhibit elevated iron concentrations and 265). The previous data error resulted from a software glitch that prevented the filter wheels from resetting after an imaging sequence (T. C. Sorensen, personal communication, 1997). Thus, when the next sequence of images was taken, the filter wheel was not in the designated position and would substitute the broadband "F" filter for one of the other five filters. Images were taken off nadir during segments of orbits 264 and 265, because the spacecraft slued as a maneuver to recover lost data. Data (6-10 wt % FeO) relative to the eastern highlands. Andre et al. [1977] and Conca and Hubbard [1979] report similar compositional differences for A1 and Mg between the eastern and western highlands bounding Smythii (Table 1). Terra materials east of Smythii are higher in A1 and lower in Mg than the those to the west, as noted in the Clementinc-derived iron maps. The east-west highland compositional division was previously attributed to mafic debris from later impact events, such as

4222 GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB 70 ø E 20 ø N 8.0 ø 90 ø 100 ø E :20 ø N NpNt NpNt Im lm l0 ø o o 10 ø 20 ø S 70 ø E 8.0 ø 90 ø 100 ø E Figure 3. Geologic map of the eastern limb region modified from Wilhehns and El-Baz [1977] Solid lines in' dicated full basin rings, and dashed lines indicate partial basin rings. See legend for description of map units.

GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB 4223 Explanation of Map units: Map Legend Cc Ec Edm Em Ic Im Idhc Ifc Im 2 Idmt Idbp Ilp Idp Im Nhit Nt Nc - Copernican crater; material of sharp-rimmed, rayed craters. - Eratosthenian Crater; material of sharp-rimmed craters. - Eratosthenian dark mantle material; Eratosthenian age pyroclastic material. - Eratosthenian mare material; Eratosthenian age basaltic material. - Imbrian crater; material of less sharp-rimmed craters. - Imbrian mare material undivided; Imbrian age basalt material, age relative to Oriental unknown. - Imbrian dark halo crater; craters with low-albedo ejecta deposits. These craters have penetrated through overlying highlands material to excavated subsurface mafic material in their ejecta. - Imbrian floor-fractured crater; characterized by a raised, fractured crater floor which may contain basalt and/or pyroclastic material. - Imbrian mare material; Imbrian age basaltic material younger than the Orientale basin. - Imbrian dark mantled terra; topographically rough, low-albedo pyroclastic material mantled over highlands terrane. - Imbrian dark basin plains; low-lying, pitted, moderate-albedo material. A mixture of volcanic material and basin floor material of Smythii. - Imbrian light plains; smooth light colored plains material. - Imbrian dark plains, smooth low-albedo material. - Imbrian mare material; Imbrian age basaltic material older than the Orientale basin. - Nectarian high-iron terra; highland material with elevated iron concentrations. - Nectarian terra; heavily cratered highlands material. - Nectarian crater; subdued-rimmed craters. NpNt - Nectarian, prenectarian undivided terra. pnc - prenectarian crater material; very subdued-rimmed crater. pncb - prenectarian corrugated basin material; smooth, angular fractured material. pnbr - prenectarian basin rim; basin massif material. pnbf - prenectarian basin floor; basin floor material. crater rim crest Crest of Basin Ring Dashed where inferred Figure 3. (continued)

:.. ß. 4224 GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB I00 ø.?: :!: '"-::4" :il.::... 80 ø 90 ø 100 ø...?,.:... 30 ø... :...::. :., '--,'- ::::.: 'J,,...... - :. ;:' ':.t:::....:: '" ;... 20 ø 10 ø 10 ø i. ;:" ' ' <':.. -'-, I :.; ' ' '. 'ø' '5,:-- -.';. ' ½-'.-. '*'"11 0 o..,½-.:½..10 ø 80 ø 90 ø -10 ø Figure 4. Topography of the Smythii-Marginis region as shown by Clementine altimetry [Zuber et al., 1994J. Each contour line represents a difference in elevation of 1000 m. The topography is merged with the shaded relief map to illustrate morphologic and topographicorrelations. The inner nm of the Smythii basin is very apparent, while a similar circular topographic high is not detected for the Marginis basin. Crisium Basin, superposed on the western side of the feldspathic Smythii basin ejecta [Andre e! al., 1977]. The C!ementinederived iron map (Plate 1) indicates, however, that the compositional dichotomy between the highlands to the east and west of Smythii is a manifestation of mare deposits presento the west of Smythii and absent east of Smythii (Figures 2 and 3). There are the range in FeO concentration along most of the basin ring is 2-6 wt %, compositionally similar to the surrounding highlands (Plate 1, Table 1). These data support the finding by Spudis' and Hood [1988] of highly feldspathic material observed in the northwest corner of the 360 km diameter ring and extend the known occurrences of anorthoatic material along most of this ring (Figure 3). The age and relatively degraded state of the innermost basin ring suggest that this anorthositic material is more likely to be comminuted soil, low in iron, rather than multiple basalt deposits scattered about the western edge of the Smythii basin. The iron map shows high-iron basalt deposits (10-14 wt % FeO) encompassed by overlapping aprons of loweriron (6-10 wt % FeO) material. Thus gardening of high-iron, blocky outcrops of pure anorthosite such as those seen at the low-aluminum basalt deposits with low-iron, high-aluminum Orientale, Humorurn, and Nectaris basins [Bussey and Spudis, highlands material has elevated the iron content at the expense of 1997; HawIce, 1993; Spudis et al., 1984, 1989]. alumina in this region. In addition to the inner basin ring of Smythii, central peaks of C!ementine data reveal anorthositic material in the innermost large craters on the eastern limb also sample material rich in anring structure of Smythii. The iron distribution map shows that orthosite. The central peaks of Joliot, Neper, and Lomonosov all

GILLIS AND SPUDIS' GEOLOGY OF THE MOON'S EASTERN LIMB 4225 Figure 5. An Apollo photograph showing the geography of Smythii Basin. Avery (A); Cam6ens (C); Dumas (D); Doyle (DO); Haldane (H); Helmeft (HE); Kiess (K); Kao (KA); K/istner (KS), K/istner G (KSG); Mcadie (M); Neper (N); Peek (P); Runge (R); Swasey (S); Widmannst/ittan (W); Warner (WA); race is placed above craters with mafic crater ejecta; Rilles and arrow depict promontory with sinuous rilles; the white box outlines the location of Plate 2. The black crosses are fiduciary marks on the Apollo Hasselblad prints (AS 15-95-12991). have compositions low in iron, 2-6 wt % FeO (Plate 1). Evidence of feldspathic material within both the inner ring of Smythii basin and its ejecta and within the central peaks of complex craters suggests that the bulk crust in this region is anortho- basin floor of Smythii affirms the protracted elapse of time between basin formation and mare emplacement. 3.4. Basalt Composition sitic between depths sampled by crater central peaks (---10 km The volcanic deposits in Smythii are 16-18 wt % FeO and 2.5- [Grieve and Garvin, 1984]) and the inner basin ring structure 3.5 wt % TiO2, as determined from the C!ementine data. These (20-30 km [Grieve et al., 1981])[Gillis et al., 1997]. data are consistent with Mariner 10 data, 4+1 wt % TiO 2 [Robinson et al., 1992] and the Apollo y-ray concentrations for FeO, 3.3. Basalt Deposits Mare basalt deposits fill the northeast interior of Smythii basin 13+2.5 wt %, and TiO2, 4+1.8 wt %, reported by Davis [1980]. The agreement in measured FeO and TiO 2 content between and form isolated lenses within some of the modified floor- C!ementine and Mariner 10 multispectral data [Robinson et al., fractured craters (Figures 3 and 5). The main mare deposit occu- 1992] and Apollo gamma-ray data [Davis, 1980] offers assurance pies an area of 32,000 km 2 or-- 25% of the basin floor surface that the techniques for calculating elemental abundaces [Lucey et area. The isolated basalt deposits have a combined areal extent of 9000 km 2. The basalts are characterized by low-albedo, highiron content, a smooth surface (with the exception of mare ridges), and low crater densities. The lack of visible domes, sinuous rilles, and flow fronts in the main mare deposit is evidence for highly effusive, flood-type eruptions [Greeley, 1976]. Crater densities for the entire main mare unit in Smythii, relative al., 1998] are valid and can be applied with confidence to the Moon globally. The Clementine data set has the added benefit of higher resolution (250 m/pixel) than the Mariner 10 (---10 km [Robinson et al., 1992]) and the Apollo y-ray data, approximately 100 to 200 km [Davis, 1980]. Resolution of this quality allows us to correlate chemical features in the FeO and TiO 2 maps with geomorphic features on the surface of the Moon (e.g., Plate 2). to the lava flows at the dated Apollo sites, indicate that these 3.5. flows are among the youngest on the Moon, 1-2 Gyr old [Schultz Basalt Thickness and Spudis, 1983; Spudis and Hood, 1988]. Age estimates by Boyce and Johnson [1978] suggesthat the deposit might be as old as 2.5+0.5 Gyr. The accumulation of premare craters on the Plate 2 illustrates that the highest iron concentrations in Mare Smythii are shifted to the north and east from the center of the mare deposit. If the observed iron content of Mare Smythii was

4226 GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB FeO Wt. ø o >8 16 1, 1 0 8 6 Plate 2. The same Apollo image as in Figure 5 reprojected and merged with the corresponding area in the C!ementine iron map (Plate 1). This image illustrates the relation of morphology (e.g., floor-fractured craters and mafic crater ejecta) with composition (e.g., low-iron crater rims and high- and low-iron ejecta deposits). White arrows point to craters with low-iron ejecta, and the yellow arrow points to the maximum-size crater with high-iron ejecta. Peek (13 km diameter) is a macroscalexample of what a crater with low-iron ejecta looks like. caused by horizontal mixing, for example, produced by crater ejecta, then the pattern of FeO concentration would mimic the boundaries of the mare unit. However, vertical mixing caused by meteorite impacts would produce a greater amount of mixing and contamination of low-iron materials in the thinner parts of the basalt unit than for the thicker areas. The vertical mixing model indicates that the iron composition of the basalt is directly related to its thickness and therefore explains why the area of highest iron content does not coincide with the center of the deposit. Thus we conclude that the basalt deposit within Smythii is thickest in the northeastern portion of the basin. The thickness of the basalt in Mare Smythii is quantified by portion Mare Smythii has forced us to supplement the previous thickness measurements using additional techniques. Accordingly, we measured the diameters of two ghost craters (2øN, 88.5øE, diameter 8 km; IøN, 89.25øE, diameter 8.75 kin) and, using equations from Pike [1974, 1977], calculated that it would require 300 and 325 m of basalt, respectively, to inundate these craters. Thus we estimate that the basalt fill of Smythii is on average 300 m thick. All thickness measurements occur outside the area of highest iron concentration within the basalt; thus the deepest portions of the basin are slightly thicker than 300 m. Our measurements of basalt thickness are consistent with the Stewart et al. [I 975] calculation of an average mare thickness of 350 m bracketing the depths at which craters have excavated highlands and nqaxinqum thickness of 450-475 m. In addition, De Hon and material or not. The composition of ejecta blankets was deter- Waskom [1976] suggest that the mare basalts average 230 m and mined using the C!ementine 750/950 image ratio and iron abunreach a maximum thickness of 500-600 m in the northern floor of dance maps (Plate 2). For simple craters, <20 km in diameter, the basin. typical excavation depths are on the order of one tenth their apparent diameter [Croft, 1980]. Craters that have excavated mate- 3.6. Surface features within the mare rial from a depth of-300 m or less (e.g., an unnamed crater 3.25 km in diameter located at 0.7øN, 87.5øE) have not excavated low-iron materials (e.g., their ejecta blankets have a high 750/950 ratio and wt % FeO). Craters with excavation depths >350 m (e.g., an unnamed crater 3.5 km in diameter located at 0.6øN, Remnant sinuous rilles located on a patch of high-stand basin floor material (89.25øE, 3.25øN) avoided burial by mare flooding (Figure 5). These rilles emplaced some basalt in the area, but their length and width (10 km long, 500 m wide) suggest that they were not responsible for the majority of the basalt deposited 86.5øE and one 3.7 km in diameter located at 3øN, 89.8øE) have excavated nonmare material from beneath the basalt (e.g., at least part of the crater ejecta blanket has a low 750/950 ratio and wt % FeO). The absence of C!ementine data combined with the absence of craters large enough to excavate nonmare material in central within the basin. Additional volcanic rilles are likely to have existed on the basin floor of Smythii, but they were submerged by the continued eruption of mare basalt. Wrinkle ridges within Mare Smythii trend in a northeasterly direction. The mare ridges correlate with the thickest accumulation of basalt. Their nonarcuate shape implies that they do not

GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB 4227 Table 2. Mixing Results Unit 415 nm 750 nm 950 nm FeO TiO2 A120 MgO Th Avg. Highlands 0.107 0.175 0.1991 5.0 0.1 28.5 4.5 0.7 Mare Smythii 0.070 0.112 0.1242 17.0 3.0 19.0 11.5 2.4 Mixture 50:50 0.088 0.143 0.1617 11.0 1.6 23.75 8 1.55 Dark basin plains 0.086 0.142 0.1614 9.0 1.3 23.5 10 1.4 % Difference 2.4 0.9 0.1 22 24 1.1 20 10.7 Using a two-component linear mixture of the spectrand composition lbr the highlandsurrounding the Smythii basin and the mare basalts within the basin, a hypothetical dark basin plains lithology (mixture 50:50) was calculated. The model result is compared to the spectrand composition of the observedark basin plains unit and reported as percent difference (%Difi: = (observ-calc)/observx100). A fifty-fifty mixture of the two lithologies closely matches the spectral data best, but a sixty-lbrty mixture of highlands and mare better approximates the composition of the observedark basin plains. reflect buried premare topography. Thus we attribute their formation to stresses developed during consolidation, local isostatic adjustments of the basin, or sagging of the basin floor in response to the overburden of the basalt flows. Greeley et al. [1977] proposed that the mare lavas flowed generally from the southeast and west toward the northeast and accumulated in the lowest parts of the basin. However, we observe that the northeastern area of the basin maintains the highest elevation (-4.25 km) and slopes downward to the west (-4.5 kin) and southwest (-4.4 km) below the mean datum of 1738 km (Figure 4). We infer from this observation that eruptions occurred in the northeast and flowed to the west and southwest. This hypothesis would explain how the basalt thins to the southwest while lacking any obvious volcanic surface features. We conclude from the absence of visible flow fronts, uniform spectral character, crater density, and iron and titanium concentrations of the main basalt deposi that the majority of the volcanic surface was deposited in a single eruptive event and from a single source at depth. Some areas in the western and eastern Smythii appear analogous to the corrugated Maunder Formation of the Orientale basin (Figure 6) [Head, 1974; Scott et al., 1977; McCauley, 1977]. This unit consists of smooth, angular blocks with concentric and radial fractures, has an FeO concentration of 6-8 wt %, and is situated between the inner ring and mare basalt. The areal extent is limited to the northwestern and eastern edges of the basin. This limited surface exposure is a consequence of greater degradation by crater erosion and burial of $mythii relative to Orien- (A) *.7 g Figure 6. Images comparing (a) the Corrugated Maunder (CM) and Srnooth Maunder (SM) Formation of the Orientale basin (LO lv-195-hi) and (b) the smooth angular blocks found in Mare Smythii (LO-I-19M). The black line in Figure 6a separates the Orientale melt sheet (CM and SM) from Mare Orientale (MO). The smooth texture of the blocks in Smythii basin (Sms) is similar in morphology to the Orientale Maunder Formation and by analogy could be part of the melt sheet from the Smythiimpact. An archetypal dark halo crater (dhc) is located in the upper center of the image. The crater has excavatedarker possible pyroclastic material from beneath the mare surface. A wrinkle ridge (wr) runs north-south along the left-hand side of the image. (B)

4228 GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB tale. If this smooth corrugated material is similar in origin to the Observations of craters within the moderate-albedo basin Maunder Formation, then it consists of the impact melt from the Smythii basin-forming event. Likewise, the inner ring of the Smythii basin, 360 km, would then be analogous to the 320 km inner Rook ring of the Orientale basin. An alternativ explanation is that concentric and radial fractures are fault grabens developed in response to tension within the mare basalts. Sagging of the basin interior, due to the overlying weight of basalt, produced graben structures along the edge of the basin. plains unit having mafic ejecta blankets (Figure 5; Plate 2) support the hypothesis of buried mare. The ejecta of multiple craters within the moderate-albedo unit (e.g., Avery and three unnamed craters labeled mce (Figure 5)) exhibit a mafic signature in the Clementine three-coloratio image and in the iron map (Plate 2). Furthermore, the ejecta from an unnamed crater (83.4øE, 3.5øS) was revealed by Conca and Hubbard [1979] to have an elevated Mg/Si ratio in the Apollo X-ray data. Finally, an unnamed dark halo crater is located north of Widmannst itten (85.1øE, 3.75øS); 4. Moderate Albedo Unit its irregular shape suggests that it is possibly of endogenic origin [Head and Wilson, 1979]. These observations provide evidence Apollo and Clementine images (Figures 1) of the Smythii ba- that the chemical composition of the dark plains subsurface unit sin show that moderate-albedo material occupies the southwest- is more mafic than the exposed surface layer. Additional possiern and central floor of Smythii basin (dbp in Figure 1 and Idbp ble eruptive source areas are degraded pit craters and rilles lounit in Figure 3). This geologic unit is hummocky in appearance cated within the dark basin material on the southeast margin of and is heavily cratered by small impacts (Figure 5). It is intermediate in albedo, A1/Si and Mg/Si ratios [Stewart et al., 1975; Andre et al., 1977], and iron composition between mare and adjacent highland terra (Plate 2; Table 1). Crater densities of the moderate albedo basin plains material yield an age estimate of -3.5 Gyr [Boyce and Johnson, 1978], older than the mare. Topographically higher than the main mare deposit (Figure 4), the basin between the craters Avery and Dumas and to the north of Kiess and Widmannst itten (Figure 5). We conclude that mare basalt erupted early, conformed to the inner ring of Smythii, and subsequently was overlain by highlands material ejected from nearby craters. We propose that the postbasin craters responsible for depositing highlands ejecta over the mare on the southwestern side of Smythii are Doyle, Halthe moderate albedo unit forms an elongate arch that stretches dane, Kiess, Warner, Widmannst itten, Neper, La P6rouse (10øS, from northwesto southeast and divides the main basalt deposit to the northeast from isolated basalt patches in the craters Dumas, Kiess, Widmannst itten, Helmeft, and Kao (Figures 3 and 5). On the basis of albedo and crater data, Stewart et al. [1975] concluded that the moderate albedo basin material was emplaced as a debris blanket of mixed origin and composition, derived 76øE; 70 kin), Ansgarius (13øS, 80øE; 90 kin), Sklodowska (19øS, 96øE; 115 km), and Humboldt (27øS, 81øE; 200 km). Over time, gardening of the surface by meteorite bombardment mixed the two lithologies [Quaide and Oberbeck, 1975] to produce the final basin plains material, intermediate in albedo and composition. from the Crisium basin, other large impact basins, and craters in the area. Conca and Hubbard [1979] state that the high-albedo 5. Floor-Fractured of the mantled terra unit rules out the presence of substantial Craters amounts of volcanic material (< 25%) as either pyroclastic mate- Premare craters within the boundaries of the moderate albedo rial or lava flows. There are some inconsistencies with these plains have been modified by endogenic processes which include models. First, the age of the dark basin material is late Nectarian or early lmbrian, too young to be the result of Crisium, which is mid-nectarian age [Wilhelms, 1987]. Second, it is fortuitous that the moderate albedo unit was emplaced only within the confines of Smythii basin. Other local topographic lows, such as the craters K istner and K istner G, do not display the moderate albedo seen in the dark plains unit (Figures 1, 2, and 5). Both craters are filled with light plains material with iron concentrations of only 2-8 wt %, lower than the dark basin floor material (6-11 wt % FeO, Plate 1). Finally, a fifty-fifty mixture of highlands and mare material would produce the observed reflectance values for the dark basin material (Table 2). This same mixture of highlands and mare would also yield the observed iron and titanium concentrations. Similar values for average A!/Si ratios [Andre et al., 1977] and FeO concentrations for the highlands west of Smythii (Plate 1), for portions of A1-Khwarizmi/King [Clark and Hawke, 1991], and for material in the unflooded portion of Smythii basin suggest that these deposits all formed by similar processes. Conuplifted and fractured floors, pyroclastic deposits, sinuous rilles, and basalt deposits. The crater rims, central peaks, and uplifted annula rings of these floor-fractured craters are lower in iron (4-8 wt % FeO) than the mare and basin plains material surrounding them (Plate 2). This indicates that these craters formed on the surface of the ancient basin floor. Two models have been proposed to explain the processes that have modified these crater floors: volcanism and plutonism [Brennan, 1975; Schultz, 1976b] and the viscous relaxation of crater topography [Baldwin, 1968; Hall et al., 1981]. The first model employs dike-fed, basaltic intrusions into the fractured material beneath the impact crater. As intrusions persist, the crater floor elevates and fractures. Along the deepest fractures, volcanic eruptions may occur. In contrast, the viscous relaxation model incorporates crustal heating to locally lower viscosity, allowing relaxation of crater topography. Mantle uplift is the mechanism for the relatively near-surface crustal heating that provides the thermal conditions for both viscous relaxation and volcanism. Studies of the floor-fractured craters in Smythii basin have led ceivably, basaltic material mixed with highland-type lithologies previous workers [Hall et al., 1981; Yingst and Head, 1998] to to account for the intermediate albedo and composition between mare and highlands. However, Conca and Hubbard [1979] noted the difficulty with lateral mixing of two lithologies. Mixing would be more efficient if it were vertical [Rhodes, 1977], suggesthat crater floor modification occurred in response to viscous relaxation of the crater. However, our observational data, as well as geophysical modeling [Dombard and Gillis, 1999], reveal inconsistencies in the viscous relaxation model. either as a thin basalt deposit reworked with basin material or as Floor-fractured craters and unmodified craters of similar size and basalt covered by ejecta from nearby craters [Schultz and Spudis, 1979, 1983]. age occur side by side; in other regions, relatively small craters have undergone greater modification and floor fracturing than

GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB 4229,100 ø 30 ' : 11,'i!::., ß 2-,.: '..,?.-: 80 ø 90 ø -I00 ø?... ::...i i -. 30 ø 20 ø,10 ø lo ø.,!0 o 80 ø 90 ø Figure 7. Bouguer gravity map [Zuber et al., 1994] merged with a shaded relief image of the eastern limb region of the Moon. Each contourepresents 100 mgals. The greatest positive Bouguer signature does not correlate with the thickest area of basalt within the Smythii basin; the highest gravity is offseto the southwest. The Marginis basin displays only a moderate positive Bouguer anomaly. Gravity highs are also associated with the crater Neper and an area northwest of Marginis Basin. adjacent larger craters [Schultz, 1976b]. Moreover, comparing a gravity anomaly (Figures 5 and 7), as required by the viscous map of the global distribution of floor-fractured craters [Schultz, relaxation model [Brennan, 1975]. In addition, calculated rates 1976b] with mare basalt distribution, basin location and gravity for deformation [Yingst and Head, 1998] suggest that craters data show [Lemoin et al., 1997] that a number of floor-fractured closest to the major mare deposit and largest gravity anomaly craters occur at significant distances from the nearest mare unit, have taken longer to modify than craters of the same diameter impact basin, or gravity anomaly. The viscous relaxation model farther from the area of greatesthermal activity. For example, requires craters with fractured floors to be proximal to thermal Runge (38 km diameter) has taken almost 5 times longer to deanomalies. However, crater floor modification by subsurface form than the crater Haldane (39 km diameter), 3.5x106 versus magma emplacement is less dependent than crustal heating to 1.6x10 * years (a difference of 12.5 million years) [Hall et al., nearby thermal anomalies because isolated mare units occur 1981 ]. Even Dumas (17 km diameter), located farthest from the within even the thickest crust of the farside (e.g., Kohlsch'dtter main mare deposit, deformed more quickly than Runge (l.2x10* and Lacus Solitudinis), far removed from any major body of years [Hall et al., 1981]). The viscous relaxation model also mare basalt. requires that the viscosity of the lithosphere vary significantly Within the Smythii basin, all of the floor-fractured craters oc- over short horizontal distances. This is the case for Ritter (28 km cur adjacento the mare and do not coincide with the highest in diameter) and Sabine (30 km diameter) craters located in

4230 GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB southwestern Tranquillitatis. These craters are equally deformed. were produced at approximately the same time, and are located within the same basalt strata, and their rims are separated by only 4 kin, but it took 4.4x107 years longer for the modification of Sabine crater. A variation in viscosity of this magnitude across such small distances seems geologically implausible. The fact that fractures do not extend outside the crater rims suggests that the mechanism for formation is near the surface and localized beneath the crater. Moreover, the process has to be able to account for deformation of large craters (e.g., Humboldt 209 km [Baldwin, 1968]) as well as small craters (e.g., unnamed craters in Aitken, 6 km in diameter [EI-Baz, 1973; Bryan and Adams, 1974]). It is counterintuitive that the deformation caused by mantle upwellings could be exclusively contained within a crater with a radius of 6 kin. Furthermore, viscosities used to model viscous relaxation are inferred to be on the order of the Earth's mantle (1022 poise [Hall et al., 1981]). In order for the floor of the craters to fracture, they must be brittle; the low theologies of the near-surface lithosphere, required by the viscous relaxation model, are contradictory to this fact. The Smythii basin contains at least 11 floor-fractured craters, the densities of which suggesthat this region has had an intense amount of subsurface igneous activity, consistent with our interpretation of the origin of the moderate albedo unit discussed in section 4. The collection of floor-fractured craters illustrates a continuum of modification (Figure 5); from craters with welldeveloped annular rings and little mare (Haldane, Doyle, Mcadie, and Warner) to those that are more enveloped by basalt (Runge, Cam6ens, Kiess, Widmannst itten, Helmert, Kao, and Swasey). Pyroelastic material is associated with irregular pits and rilles found within and adjacent to the floor-fractured craters. The irregular pits within the floor-fractured craters suggest that a near-surface void was present, foundering the overlying surface and allowing lava to drain back into the subsurface. Floorfracture craters such as Haldane, Runge, Swasey, Warner, and Widmannst itten contain pyroelastic material within their crater walls. The floor-fractured craters Doyle, Cam6ens, Tasso, and Keiss have pyroelastic material along their peripheries. The pyroclastic deposits and sinuous rilles indicate that highly effusive, mare flood-type volcanism occurred in the central portion of the basin, while fire-fountaining, plains-type eruptions occurred adjacento the main mare deposit. 6. Marginis Basin 6.1. Regional Geology Mare Marginis (20øN, 84øE) may or may not occupy an impact basin. The Marginis basin has been mapped as a pre- NectarJan impact basin, 580 km in diameter [El-Baz and Wilhelms, 1975; I4/ilhelms, 1987]. However, Clementine altimctry shows only a shallow, irregular depression lacking any topographic rim (Figure 4). Moreover, there is no indication of an anorthositic ring surrounding the basin in Clementine images (Plate 1). Marginis also lacks a strong (>200 mgals), circular positive gravity anomaly (Figure 7) displayed by many large impact basins. These attributes are characteristic of a very old, highly degraded impact basin, a coalescence of large craters, or a structural trough exterior and concentric to Smythii basin igillis et al., 1997]. 6.2. Basalt Deposits The iron and titanium abundances for the basalt in Mare Mar- ginis are similar to those in Mare Smythii: 15-17 wt % FeO; 2.6-3.6 wt % TiO 2 (Plate 1; Table 1). Similar titanium concentrations have been reported by Robinson et al. [1992] for Mare Marginis (3+1 wt % TiO2) and Mare Smythii (4+1 wt % TiO2). The extent of mare is 50,300 km 2, slightly larger than Mare Smythii. Mare Marginis is older than Mare Smythii, as indicated by higher crater densities [Boyce and Johnson, 1978]. There are two distinct episodes of basaltic volcanism within the basin: Im and lm 2 in Figure 3. The chemical compositions of the basalt flows are similar (Plate 1), suggesting that they originated from a related source region. The two units are distinguished by their different crater densities. The younger of the two units covers most of the surface of Mare Marginis. The older unit is present along the eastern highland-mare boundary of the basin and along the south central portion of the basin, north of the craters Neper and Jansky. 6.3. Basalt Thickness The absence of craters larger than 2 km within the central portion of Mare Marginis prevents us from using the method described in section 3.5 for estimating basalt thickness; craters with ejecta consisting of continuous, low-iron material are not observed. Thus a second method of calculating mare thickness from measurements of the exposed rim height of partially buried craters [Eggleton et al., 1974; De Hon and Waskom, 1976; De Hon, 1979] was used to estimate the maximum thickness of the basalt. The mare thickness around three, unnamed, partially buried craters (Plate 1; 13øN, 87.8øE, 12.5 km diameter; 13øN, 88.7øE, 12.3 km diameter; and 13.2øN, 83.7øE, 11.25 km diameter) range between 150 and 320 m. Therefore we suggest that the average basalt thickness in Marginis basin is 250 m. Moreover, we conclude that the thickest portion of the basalt corresponds with the region of highest iron content, the southwest central portion of Mare Marginis (Plate i), on the basis that vertical mixing has been shown to dominate over horizontal mixing [Rhodes, 1977]. 6.4. Volcanic Surface Features The absence of volcanic domes, flow fronts, and sinuous rilles in Mare Marginis suggests that these basalts were the result of highly effusive, flood-type eruptions like those in Mare Smythii. The only noted exception is a degraded ri!le located atop a promontory (86.3øE, 13.8øN) (Plate I). Mare Marginis also has no apparent wrinkle ridges. This fact indicates that the basin-fill has not been tectonically compensated subsequent to basalt flooding like most mare-filled impact basins. 7. Lunar Swirl Material Unusual swirl patterns of alternating bright and dark material occur along the highland-mare boundary of northern Marginis basin (Figures 1, and 2)[EI-Baz, 1972]. Associated with the swirls are strong magnetic anomalies detected by the Apollo subsatellite magnetometer [Srnka and Schultz, 1979; 1tood and Williams, 1989]. The swirls have been suggested to be either a chemical alteration of surface materials [Schultz, 1976a], deposition of sublimates degassed from the interior of the Moon, scouring of the surface by recent cometary impacts [Srnka and Schultz, 1979; Schultz and Srnka, 1980], or agglutinate-free regolith that has been shielded from the ion bombardment of the solar wind by the associated anomalies of high surface magnetism [Hood and Williams, 1989]. Using the Clementine data, we can support a swirl generation mechanism that produces an agglutinate-poor soil from in situ

GILLIS AND SPUDIS' GEOLOGY OF THE MOON'S EASTERN LIMB 4231 0.325 0.3 0.275 0.25 0.225 Comparison of reflectance data for swirl and non-swirl materials 0.2 0.175 0.15 0.125 O Highlands Material Highland Swirls ß []- Mare Swirls Mare material O. 1 - - - - 0.075 0.05 ' t ' I ' I ' t ' I ' 400 500 600 700 800 900 1000 Wavelength (rim) Figure 8. A comparison between reflectance values of swirl and nonswirl material, on both highlands and mare lithologies. Data were collected from the 415,750, and 950 nm wavelengths from the Clementine data set. Symbol size approximates the error. material. Both the scouring of the surface by a recent cometary impact [Srnka and Schultz, 1979; Schultz and Srnka, 1980] and ionic solar wind shielding by associated magnetic anomalies [Hood and Williams, 1989] are consistent with this mode of formation. Clementinc band ratios (415/750 and 750/950) yield higher values for the swirl material than the nonswirl material; similar high values are recorded for the bright Copernican crater rays of Giordano Bruno. Figure 8 shows that the regolith superposed by swirl material is -62% higher in reflectivity than the same unit without swirl material. Mare basalt within the swirl gravity readings occur to the south and east of the main mare deposit and do not correspond with the thickest deposit of basalt as reported by Brennan [1975] and EI-Baz and Wilhelms [1975]. Furthermore, a mason of similar magnitude should be associated with Mare Marginis because of its comparable thickness to Mare Smythii. Instead, there exists only a moderate (-125 mgals) mascon beneath the basalt flows of Mare Marginis. The basalt flows in Smythii Basin are too thin (-300 m) to be the primary cause of the mascon; it requires a 18 km thick layer of material with a density of 3.3 g/cm 3. Similar associations of large positive Bouguer gravity and thin basaltic fill are noted for the Nectaris and Orientale basins [Head, 1974, 1976]. Therefore we conclude that a submare mass contribution is required to match the observed positive Bouguer gravity data over Smythii. The existence of the submare mass is the result of the excavation of relatively low density crustal material (2.8 g/cm 3) by the impact event, followed by a rising plug of greater density mantle (3.3 g/cm 3) to compensate isostatically for the excavated crustal material [Wise and Yates, 1970; Phillips and Dvorak, 1981]. 9. Geologic Evolution of the Eastern Limb Region Iron values for the central peaks of large craters and for the inner ring and ejecta of the Smythii basin show that the bulk composition of the eastern limb highlands is uniform both laterally and vertically. Therefore elevated levels of iron and diminished levels of aluminum along the western boundary of the Smythii basin cannot be explained by an impact into the chemical heterogeneous target. Instead, basalt deposits to the west of Smythii basin have been mixed with surrounding highlands by meteorite churning of the lunar regolith. Mare Marginis has undergone multiple episodes of volcanism, from early Imbrian to mid-eratosthenian [Boyce and Johnson, 1978]. The difference in age between the different units is detected by the greater crater density of one unit over the other. The chemical compositions of the flows are similar, which suggests that they originated from a similar source material. Also, the lack of volcanic rilles, domes, and flow fronts suggests that both eruptive events occurred as flood-type volcanism. The mare basalt within the Smythii basin is among the youngest on the Moon. There are two mechanisms for the emplacement of basalt into the Smythii basin. Plains-type volcanism is noted by the occurrence of rilles and pyroclastic deposits found among elevated topography and within floor-fractured craters. The smooth surface of the main mare deposit indicates these units were deposited by highly effusive, flood-type volcanism. The early stages of basin filling were low-volume, plains-type material has a relatively flatter or "bluet" continuum slope and deeper 950 nm absorption band than mare basalt outside the swirl material (Figure 8). Similar observations were made by Bell and Hawke [1981] for the Reiner Gamma Formation and were interpreted as signs of an immature regolith, free of space weathering effects. In addition, the iron map (Plate 1) shows that the swirl material has the same composition as the background material they lie upon (the method for calculating iron abundance [Lucey et al., 1998; Blewett et al., 1997] decouples the effects caused by albedo and soil maturity differences). These findings are consistent with data from Adams and McCord [1971], who compared Apollo 12 surface samples with core tube samples taken from 20 cm below the surface. Material from 20 cm depth had a 40% higher reflectivity than the surface fines. Adams and McCord [1973] also observed a negative correlation between agglutinate content of Apollo 16 soils and albedo. volcanism, and later eruptive emplacements were flood basalts. Basaltic material was not emplaced to an approximate hydrostatic level during mare volcanism in the eastern limb region. In contrast, ToksOz and Solomon [1973] suggesthat thickening of the lunar lithosphere with time resulted in an increase in source 8. East Limb Gravity depth; thus basins filled to successively higher levels [Lucchitta In both the Clementinc [Zuber et al., 1994] and Lunar Prospector [Konopliv et al., 1998] gravity data, Smythii basin exhibits a strong (>500 mgals) Bouguer gravity anomaly (Figure 7). and Boyce, 1979]. In the Marginis-Smythii region, basalts have not been emplaced to a common topographic level. El-Baz and Wilhelms [1975] noted a deposit at 73øE, 9øN, outside of Mare The subsurface concentration of mass or "mascon" beneath Smythii, that was 3 km higher in elevation than the Smythii de- Smythii is asymmetric and displaced toward the eastern part of the basin. If this mascon was the sole consequence of the mare basalt, then the greatest gravity readings would correlate with the posits. Even within Smythii alone, basalts do not reach an average hydrostatic level. The main mare deposit lies between -4.6 and -4.2 km below the mean lunar radius (Figure 4). Basalt in thickest basalt deposits in Mare Smythii. Instead, the greatest the craters Kiess, Widmannst itten, Kao, and Helmeft is only -2.8

4.232 GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB km below the mean datum. Thus we conclude that there is no correlation between age of volcanic deposits and their elevation for eastern limb of the Moon. There is a difference of 2 km between the surface level of young basalt in Mare Smythii (-4.4 kin) and the older basalts in Mare Marginis (-2.55 km). Thus premare topography is the main factor controlling the level of mare basalt emplacement. Basins with thin basalt flows and large positive gravity anomalies (i.e., Smythii, Nectaris, and Orientale) are evidence that mascons are primarily the result of relief on the lunar mantle. Basins that have comparable basalt thicknesses and do not display mascons of similar magnitude (e.g., Marginis) conceivably lack the upwelled mantle either because they are not the direct result of an impact or have been isostatically compensated. The absence of wrinkle ridges within Mare Marginis is evidence for minimal isostatic adjustment. 10. Conclusions The bulk crust of the eastern limb of the Moon is anorthositic between depths sampled by the central peaks of complex craters, -10 km (e.g., Joliot, Neper, and Lomonosov), and the inner basin ring structure of Smythii,-30 km. The compositional dichotomy observed for MgO, A1203, and FeO between the eastern-western portion of Smythii basin is the result of regolith mixing between mare and highlands lithologies. The presence of basalt deposits west of the Smythii basin has increased the FeO and MgO content of the soil at the expense of AI20 3. Marginis basin is not an impact basin. Its structure, composition, and gravity are more consistent with a coalescence of large craters or a structural trough exterior and concentric to the Smythii basin. The basalt deposits in Marginis basin are similar in composition to those of Mare Smythii. We estimate an average thickness of 300 m for the basalt in Smythii and 250 m in Marginis. On the basis of crater density, at least two eruptive events occurred within the Marginis basin. Both periods of volcanic activity are older than the volcanic deposits in Mare Smythii. The dark basin plains formation within Smythii basin was produced by an approximately fifty-fifty mixture of highlands and basaltic material. NectarJan or early lmbrian age basaltic lava erupted onto the basin floor of Smythii and conformed to the shape of the inner ring of Smythii. Over time, gardening of the lunar surface by meteorite impacts mixed the two lithologies and produced the basin plains unit that is intermediate in albedo and chemistry. Examining the morphologic and geologic character of floorfracture craters within the Smythii basin reveals basalt deposits, pyroclastic material, sinuous rilles, and deformation occurring on a very localized, independent, and brief timescale. We conclude on the basis of this evidence that the most sound explanation for the origin of floor-fracture craters is dike-fed basaltic intrusions beneath a crater floor. Clementine multispectral data and FeO maps show that swirl material is agglutinate poor, optically fresh in situ material with the same composition as the material it overlies. These findings support either a recent cometary impact scouring away the spaceweathered surface regolith or the interdiction of agglutinate formation by magnetic shielding of the ionic solar wind as models for swirl formation. Strong positive Bouguer gravity anomalies associated with thin mare fill in the Smythii basin supporthe model of Wise and Yates [1970] that concentrations of mass are the primary conse- quence of an uplift of denser mantle material in response isostatic adjustments as a result of crater excavation. Acknowledgments. We would like to thank Ben Bussey, John Guest, and internal reviewers Brad JollitY, Randy Korotev, and Ryan Zeigler for their insightful comments and suggestions that helped to improve this manuscript. This paper is LPI contribution 985. This research was supported in part by the National Aeronautics and Space Administration Graduate Research Program and grant NAG5-6784 (BLJ). References Adams, J. B., and T. B. McCord, Alteration of lunar optical properties: Age and composition effects, Science, 171,567-571, 1971. Adams, J. B., and T. B. McCord, Vitrification darkening in the lunar highlands and identification of Decartes material at the Apollo 16 site, Proc. Lunar Sci. Conf. 4th, 163-177, 1973. Andre, C. G., R. W. Wolfe, I. Adler, P. E. Clark, J. R. Weidner, and J. A. Philpotts, Chemical character of the partially flooded Smythii basin on A1/Si orbital x-ray data, Proc. Lunar Sci. Conf. 8th, 925-931, 1977. Baldwin, R. B., Rille pattern in the lunar crater Humboldt, J. Geophys: Res., 73, 3227-3229, 1968. Bell, J. F., and B. R. Hawke, The Reiner Gamma Formation: Composition and origin as derived from remote sensing observations, Proc. Lunar Planet. Sci. Conf., 12B, 679-694, 1981. Binder, A. B., Lunar Prospector: Overview, Science, 281, 1475-1476, 1998. Blewett, D. T., P. G. Lucey, B. R. Hawke, and B. L. Jolliff, Clementine images of the lunar sample-return stations: Refinement of FeO and TiO2 mapping techniques, d. Geophys. Res., 102(E7), 16,319-16,325, 1997. Boyce, J. M., and D. A. Johnson, Ages of flow units in the lhr eastern maria and implications for basin-filling history, Proc. Lunar Planet. Sci Conf 9th, 3275-3283, 1978. Brennm, W. J., Evolution of Mare Smythii (abstract), Lunar Sci., VI, 86-88, 1975. Bryan, W. B., and M.-L. Adams, An intecretation of volcanic and structural features of crater Aitken, Proc. Lunar Planet. Sci Conf 5th, 25-34, 1974. Bussey, D. B. J., and P. D. Spudis, Compositional analysis of the Orientale basin using tull resolution Clementine data: Some preliminary results, Geophys. Res. Lett., 24, 445-448, 1997. Clark, P. E., and B. R. Hawke, The lunar thrside: The nature of highlands east of Mare Smythii, Earth Moon Planets, 53, 93-107, 1991. Conca, J., and N. Hubbard, Evidence of early volcanism in Mare Smythii, Proc. Lunar Planet. Sci. Conf loth, 1727-1737, 1979. Crotl, S. K., Cratering flow fields: Implications tbr the excavation and transient expansion stages of crater tbrmation, Proc. Lunar Planet. Sci. Conf 11th, 2347-2378, 1980. Davis, P. A., h'on and titanium on the Moon fi'om orbital gamma-ray spectrometry with implications lbr crustal evolutionary models, d. Geophys. Res., 85(B6), 3209-3224, 1980. De Hon, R. A., Thickness of the western mare basalts, Proc. Lunar Sci. Conf loth, 2935-2955, 1979. De Hon, R. A., and J. D. Waskom, Geologic structure of the eastern mare basins, Proc. Lunar Sci. Conf 7th, 2729-2746, 1976. Dombard, A. J., and J. J. Gillis, Simulating the tbrmation of lunar floor- fi'actured craters using elastoviscoplastic relaxation (abstract), in New I/iews of the Moon lh Integrated Remotely Sensed. Geophysical and Sample Datasets, Lunar and Planet. Inst., Houston, Tex., 1999. Eggleton, R. E., G. G. Schaber, and R. J. Pike, Photogeologic detection of surthces buried by mare basalts (abstract), Lunar Sci., I/, 200-202, 1974. El-Baz, F., The Alhazen to Abul Wafa swirl belt: An extensive field of light-colored, sinuous markings, in Apollo 16 Preliminary Science Report, pp.29-93 to 29-96, NASA, Washington, D.C., 1972. El-Baz, F., Aitken crater and its environs, in Apollo 17 Preliminary Science Report, pp.32-8 to 32-12, NASA, Washington, D.C., 1973. E1-Baz, F., and D. E. Wilhelms, Photogeological, geophysical, and geochemical data on the east side of the Moon, Proc. Lunar Sci. Conf 6th, 2721-2738, 1975. Eliason, E. M., Production of Digital image models with the ISIS system, Lunar Planet. Sci., XXVIII, abstract 1198, 331-332, 1997. Gaddis, L., Use of ISIS tbr processing and analysis of planetary remote

GILLIS AND SPUDIS: GEOLOGY OF THE MOON'S EASTERN LIMB 4233 sensing data (abstract), Geol. Soc. Am. Abst. Programs, 28(7), A-386, 1996. Gillis, J. J., P. D. Spudis, and D. B. J. Bussey, The geology of Smythii and Marginis basins using integrated remote sensing techniques: A look at what's around the comer (abstract), Lunar Sci., XXP7II, 419-420, 1997. Gillis, J. J., L. A. Haskin, and P. D. Spudis, An empirical calibration to calculate thorium abundance fi'om the Lunar Prospector gamma-ray data, Lunar Planet. Sci., XXX, abstract 1699, 1999. Greeley, R., Modes of ernplacement of basalt terrains and an analysis of mare volcanism in the Orientale basin, Proc. Lunar Sci. Conf. 7th, 2747-2759, 1976. Greeley R., P. H. Schultz, and C. L. Wilbur, Volcanic ILatures on the Smythii basin (abstract), Lunar Sci., P7II, 371-373, 1977. Grieve, R. A. F., and J. B. Garvin, A geometric model lbr excavation and modification at terrestrial simple impact craters, J. Geophys. Res., 89(B13), 11,561-11,572, 1984. Grieve, R. A. F., P. B. Robertson and M. R. Dence, Constraints on the lbrmation of ring impact structures, based on terrestrial data. in Multi-Ring Basins, Proc. Lunar Planet. Sci. Conf., 12A, 37-57, Lunar Planet. Inst. Houston, Tex., 1981. Hall, J.L., S.C. Solomon, and J. W. Head, Lunar floor-fi'actured craters: Evidence ibr viscous relaxation of crater topography, J. Geophys. Res., 86(B10), 9537-9552, 1981. Hawke, B. R., C. A. Peterson, P. G. Lucey, G. J. Taylor, D. T. Blewett, B. A. Campbell, C. R. Coombs, and P. D. Spudis, Remote sensing studies of the terrain northwest of Humorurn Basin, Geophys. Res. Lett., 20, 419-422, 1993. Head, J. W., Orientale multi-ringed basin interior and implications tbr the petrogenesis of lunar highland samples, Moon, 11, 327-356, 1974. Head, J. W., Lunar volcanism in space and time, Rev. Geophys., 14(2), 265-300, 1976. Head, J. W., and L. Wilson, Alphonsus-type dark-halo craters: Morphology, morphometry and eruption conditions, Proc. Lunar Planet. Sci. Conf. loth, 2861-2897, 1979. Hood, L. L., and C. R. Williams, The lunar swirls: Distribution and possible origins, Proc. Lunar Planet. Sci. Conf. 19th, 99-113, 1989. Konopliv, A. S., A. B. Binder, L. L. Hood, A. B. Kucinskas, W. L. Sjogren, and J. G. Williams, Improved gravity field of the Moon fi'orn Lunar Prospector, Science, 281, 1476-1480, 1998. La Jolla Consortium, Global maps of lunar geochemical, geophysical, and geologic variables, Proc. Lunar Sci. Conf 8th, fi'ontispiece, 25 plates, 1977. Lawrence, D. J., W. C. Feldman, B. L. BalTaclough, A. B. Binder, R. C. Elphic, S. Maurice, and D. R. Thomsen, Global elemental maps of the Moon: The Lunar Prospector gamma-ray spectrometer, Science, 281, 1484-1489, 1998. Lemoine, F. G. R., D. E. Smith, M. T. Zuber, G. A. Neuman and D. R. Rowlands, A 70th degree lunar gravity model (GLGM-2) fi'om Clementine and other tracking data, d. Geophys. Res., 102(E7) 16,339-16,359, 1997. Lucchitta, B. K., and J. M. Boyce, Altitude-age relationships of the lunar maria, Proc. Lunar Planet. Sci. Conf loth, 2957-2966, 1979. Lucey, P. G., G. J. Taylor, and E. Malaret, Abundance and distribution of iron on the Moon, Science, 268, 1150-1153, 1995. Lucey, P. G., D. T. Blewett, J. R. Johnson, G.J. Taylor, and B. R. Hawke, Lunar titanium content from UV-VIS measurements (abstract), Lunar Planet. Sci., XXP'II, 781-782, 1996. Lucey, P. G., D. T. Blewett, and B. R. Hawke, Mapping the FeO and TiO2 content of the lunar surihce with multispectral imagery, d. Geophys. Res., 103(E2), 3679-3700, 1998. McCauley, J. F., Orientale and Caloris, Phys. Earth Planet. Inter., 15, 220-250, 1977. McEwen, A. S., A precise lunar photometric function (abstract), Lunar Planet. Sci., XXP'II, 841-842, 1996. McEwen, A. S., E. Eliason, P. G. Lucey, E. Malaret, C. M. Pieters, M. S. Robinson, and T. Sucharski, Summary of radiometric calibration and photometric normalization steps ibr the Clementine UVVIS images, Lunar Planet. Sci., XXIX, abstract 1466, 1998. Nozette, S., et al., The Clementine Mission to the Moon: Scientific overview, Science, 266, 1835-1839, 1994. Phillips, R. J., and J. Dvorak, The origin of lunar mascons: Analysis of the Bouguer gravity associated with Grimaldi. In Multi-ring Basins, Proc. Lunar Planet. Sci. Conf, 12A, 91-104, Lunar Planet. Inst. Houston, Tex., 1981. Pike, R. J., Depth/Diameter relations of fi'esh lunar craters: Revision fi'om spacecrall data, Geophys. Res. Lett., 1,291-294, 1974. Pike, R. J., Size-dependence in the shape of fi'esh impact craters on the Moon, in Impact and Explosion Cratering, edited by D. J. Roddy, R. O. Pepin, and R. B. Merrill, pp. 489-509, Pergamon, Tarrytown, N. Y., 1977. Quaide, W., and V. Oberbeck, Development of the mare regolith: Some model considerations, Moon, 13, 27-55, 1975. Rhodes, J. M., Some compositional aspects of lunar regolith evolution, Philos. Trans. R. Soc. London, Set. A, 285, 293-301, 1977. Robinson, M. S., B. R. Hawke, P. G. Lucey, and G. A. Smith, Mariner 10 multispectral images of the eastern limb and lhrside of the Moon, d. Geophys. Res., 97(E11), 18,265-18,274, 1992. Schultz, P. H., Moon Morphology, 626 pp., Univ. of Tex. Press, Austin, 1976a. Schultz, P. H., Floor-fi'acture lunar craters, Moon, 15, 241-273, 1976b. Schultz, P. H., and P. D. Spudis, Evidence lbr ancient mare volcanism, Proc. Lunar Planet. Sci. Conf loth, 2899-2918, 1979. Schultz, P. H., and P. D. Spudis, Beginning and end of lunar mare volcanism, Nature, 302,233-236, 1983. Schultz, P. H., and L. J. Srnka, Cometary collisions on the Moon and Mercury, Nature, 284, 22-26, 1980. Scott, D. H., J. F. McCauley, and M. N. West, Geologic map of the west side of the Moon, U.S. Geol. Surv., Map, 1-1034, 1977. Spudis, P. D., The Geology of Multi-Ring Impact Basins, 263 pp., Cambridge Univ. Press, New York, 1993. Spudis, P. D., and L. L. Hood, Geological and geophysical field investigations from a lunar base at Mare Smythii, paper presented at Second ConiCfence on Lunar Bases and space Activities of the 21 century, Lyndon B. Johnson Space Center and the Lunar and Planetary Institute, Houston, Tex., 163-174, 1988. Spudis, P. D., B. R. Hawke, and P. G. Lucey, Composition of Orientale Basin deposits and implications Ibr the lunar basin-ibrming process, Proc. Lunar Planet. Sci. Conf 15th, Part 1, d. Geophys. Res., 89, suppl., C 197-C210, 1984. Spudis, P. D., B. R. Hawke, and P. G. Lucey, Geology and deposits of the lunar Nectaris basin, Proc. Lunar Planet. Sci. Conf 19th, 51-59, 1989. Srnka, L. J., and P. H. Schultz, A cometary origin of Reiner-¾ magnetic anomalies (abstract). Lunar Planet. Sci., X, 1076-1078, 1979. Stewart, H. E., J. D. Waskom, and R. A. DeHon, Photogeology and basin configuration of Mare Smythii, Proc. Lunar Sci. Conf 6th, 2541-2551, 1975. Strain, P. L., and F. E1-Baz, Smythii basin topography and comparisons with Orientale, Proc. Lunar Planet. Sci. Conf loth, 2609-2621, 1979. Toks6z, M. N., and S.C. Solomon, Thermal history and evolution of the Moon, Moon, 7, 251-278, 1973. Wilhelms, D. E., The geologic history of the Moon, U.S. Geol. Surv. Prof Pap., 1348, 1987. Wilhelms, D. E., and F. El-Baz, Geologic map of the east side of the Moon, U.S. Geol. Surv., Map, 1-948, 1977. Wise, D. U., and M. T. Yates, Mascons as structural relief on a lunar moho, d. Geophys. Res., 75(2), 261-268, 1970. Yingst, R. A., and J. W. Head, Characteristics of lunar mare deposits in Smythii and Marginis basins: Implication ibr magma transport, d. Geophys. Res., 103, 11,135-11,158, 1998. Zuber, M T., D. E. Smith, F. G. Lemoine, and G. A. Neumann, The shape and internal structure of the Moon i?om the Clementine Mission, Science, 266, 1839-1843, 1994. J.J. Gillis, Department of Earth and Planetary Sciences, Washington University, Campus Box 1169. One Brookings Dr., St. Louis, MO 63130. (gillis levee.wustl.edu) P. D. Spudis. Lunar and Planetary. Institute. Houston, TX 77058. (Received June 9, 1999; revised October 4. 1999; accepted October 12, 1999.)