Remote sensing and geologic studies of the Balmer-Kapteyn region of the Moon

Transcription

1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 110,, doi: /2004je002383, 2005 Remote sensing and geologic studies of the Balmer-Kapteyn region of the Moon B. Ray Hawke, 1 J. J. Gillis, 1 T. A. Giguere, 1,2 D. T. Blewett, 3 D. J. Lawrence, 4 P. G. Lucey, 1 G. A. Smith, 1 P. D. Spudis, 5 and G. Jeffrey Taylor 1 Received 23 November 2004; revised 22 February 2005; accepted 1 March 2005; published 16 June [1] The Balmer-Kapteyn (B-K) region is located just east of Mare Fecunditatis on the east limb of the Moon. It is centered on the Balmer-Kapteyn basin, a pre-nectarian impact structure that exhibits two rings, approximately 225 km and 450 km in diameter. Clementine multispectral images and Lunar Prospector (LP) gamma-ray spectrometer (GRS) data were used to investigate the composition, age, and origin of geologic units in the region. A major expanse of cryptomare was mapped within the B-K basin. Spectral and chemical data obtained for dark-haloed craters (DHCs) established that these impact craters excavated mare basalt from beneath higher-albedo, highland-rich surface units. The buried basalts exposed by DHCs in the region are dominated by low-titanium mare basalts. The fresh DHC FeO values ( wt.%) that best represent those of buried mare basalts are well within the range of values exhibited by high-alumina mare basalts. While most cryptomare deposits occur beneath surfaces that range in age from Imbrian to Nectarian, it is possible that some mare flows were emplaced during pre-nectarian time. Most cryptomare deposits in the B-K region were formed by the contamination of mare surfaces by highland-rich distal ejecta from surrounding impact craters. These Balmer-type cryptomare deposits are usually associated with light plains units. Major LP-GRS FeO enhancements are associated with cryptomaria in the Balmer-Kapteyn, Lomonosov-Fleming, Schiller-Schickard, and Mendel-Rydberg regions. Citation: Hawke, B. R., J. J. Gillis, T. A. Giguere, D. T. Blewett, D. J. Lawrence, P. G. Lucey, G. A. Smith, P. D. Spudis, and G. J. Taylor (2005), Remote sensing and geologic studies of the Balmer-Kapteyn region of the Moon, J. Geophys. Res., 110,, doi: /2004je Introduction and Background [2] The Balmer-Kapteyn (B-K) region is located just east of Mare Fecunditatis on the eastern limb of the Moon (Figures 1 and 2). The region is southwest of Mare Smythii, contains the craters Langrenus, Petavius, Kapteyn, and La Pérouse, and exhibits extensive light plains deposits (Figure 2). The B-K region is centered on the Balmer- Kapteyn basin, which is a pre-nectarian impact structure that exhibits two rings, approximately 225 km and 450 km in diameter. This impact basin played a major role in forming the topography of the region. [3] While early workers recognized a large impact structure in the region [Hartmann and Wood, 1971; Schaber et al., 1977], Maxwell and Andre [1981] first provided a detailed description of the ring structure. They noted that 1 Hawaii Institute of Geophysics and Planetology, University of Hawaii, Honolulu, Hawaii, USA. 2 Also at Intergraph Corporation, Honolulu, Hawaii, USA. 3 NovaSol, Honolulu, Hawaii, USA. 4 Los Alamos National Laboratory, Los Alamos, New Mexico, USA. 5 Applied Physics Laboratory, Johns Hopkins University, Laurel, Maryland, USA. Copyright 2005 by the American Geophysical Union /05/2004JE the inner ring, 225 km in diameter, is centered at 15 S and 70 E. This ring was identified on the basis of isolated rugged mountains of pre-nectarian terra that occur on the NE and NW rims of Balmer crater, and on the southwest rim of Kapteyn crater. The identification of the outer B-K ring, approximately 450 km in diameter, was made on the basis of irregularities in the rim crests of surrounding large craters, and short segments of rugged highlands between these craters [Maxwell and Andre, 1981]. The B-K basin rings were also mapped by Wilhelms [1987]. He suggested, however, that the outer ring was 550 km in diameter. [4] Geologic units within the Balmer-Kapteyn region consist of pre-nectarian to Nectarian terra, crater materials of various ages, extensive light plains deposits, and a small number of mare units of limited spatial extent [Wilhelms and El-Baz, 1977; Maxwell and Andre, 1981; Wilhelms, 1987]. An older Imbrian- or Nectarian-age light plains (INp) unit occurs predominantly between the inner and outer basin rings, and a younger, Imbrian-age light plains (Ip) unit is present within the inner ring. Wilhelms and El-Baz [1977] included these light plains units as part of a young plains geologic province that may have been produced by an episode of non-mare volcanism. Support for this hypothesis was provided by Haines et al. [1978]. These workers used the Apollo orbital ganma-ray data to identify an area with anomalously high Th concentrations (4.0 ppm) centered 1of16

2 Figure 1. Full-Moon photograph showing the location of the Balmer-Kapteyn region on the east limb. just north of Balmer crater which correlated with Imbrianage plains deposits. Haines et al. [1978] concluded that the most likely mechanism for the formation of these light plains was an episode of highland volcanism just following the termination of intense surface bombardment. [5] An alternate explanation has been proposed by several workers [e.g., Schultz and Spudis, 1979, 1983; Hawke and Spudis, 1980; Hawke et al., 1985]. Schultz and Spudis [1979, 1983] identified several dark-haloed impact craters in the Balmer-Kapteyn region and suggested that the basin was the site of ancient mare volcanism. Buried mare deposits such as these were designated as cryptomare by Head and Wilson [1992], who stated that cryptomare means covered or hidden mare deposits that are obscured from view by the emplacement of subsequent deposits of higher albedo. Hawke and Spudis [1980] and Hawke et al. [1985] identified additional dark-haloed impact craters in the B-K basin and used the Apollo orbital chemistry data sets to locate and investigate geochemical anomalies in the region. They noted that light plains deposits in the B-K region exhibited higher Th, FeO, and MgO values than those associated with the surrounding highlands. They presented arguments based on the orbital geochemistry data and the distribution of dark-haloed craters that the region was the site of early basaltic volcanism and that the resulting mare deposits were later covered by a thin higher-albedo surface layer enriched in highland debris contributed by surrounding large craters. [6] Maxwell and Andre [1981], Andre and Strain [1983], Clark and Hawke [1987], and Antonenko et al. [1995] used Apollo orbital X-ray fluorescence data to investigate the composition of surface units in the B-K region. Unfortunately, Apollo X-ray data was obtained for only the northern portion of the region. Still, in areas where coverage exists, the light plains units exhibit relatively low abundances of Al 2 O 3 and relatively high concentrations of MgO. Maxwell and Andre [1981] and Andre and Strain [1983] summarized existing orbital geochemical data for the region and presented Al/Si, Mg/Si, and Mg/Al data for the region. They noted that the Imbrian-age plains unit (Ip) exhibits Mg/Al concentration ratios greater than 0.58 near the rimless crater Kapteyn C. The older Imbrian- or Nectarian-age plains (INp) unit east of Kapteyn crater is characterized by Mg/ Al concentration ratios greater than These values are only slightly less than those determined for mare surfaces on the lunar nearside [Maxwell and Andre, 1981]. [7] Investigations of cryptomaria can provide information that is critical to understanding the evolution of the Moon. At present, the time of the onset of mare volcanism is not accurately known. Since cryptomaria represent the earliest mare basalts, determination of their ages will provide information concerning the initiation of extrusive mare volcanism. Cryptomaria were formed by magmas generated by very early partial melting of the lunar mantle. Chemical data for cryptomaria provide evidence concerning the composition of these early partial melts. [8] In the immediate post-apollo era, several investigators determined the age and composition of mare units as a function of position on the lunar surface [e.g., Soderblom and Lebofsky, 1972; Head, 1976; Soderblom et al., 1977]. They noted that basaltic deposits on the eastern portion of the lunar nearside were generally old ( Ga) and titanium-rich, while mare units on the western side were younger and exhibited relatively low TiO 2 values. These observations lead to the suggestion that the earliest mare Figure 2. Oblique view of the Balmer-Kapteyn region. The image was obtained by the MIC on board Nozomi/ JAXA. Extensive light plains occur in and around Balmer crater. These light plains deposits are surrounded by several major impact craters (i.e., Langrenus, La Pérouse, Ansgarius, Humboldt, and Petavius). 2of16

3 basalts were titanium-rich. Later, Andre et al. [1979] used the Apollo orbital X-ray data to determine that early high- MgO basalts systematically occur along mare/highland boundaries in the six lunar nearside basins for which there is orbital X-ray coverage. They proposed that the earliest stages of volcanism in these basins emplaced high-magnesium mare basalts and that these early flows were buried by younger, less-magnesian basalts. Andre et al. [1979] concluded that the earliest mare basalts were magnesium-rich and widespread on the Moon. Since it is now known that cryptomaria are the oldest mare basalts [e.g., Head and Wilson, 1992], determination of the compositions of the buried mare units will allow these hypotheses to be tested. [9] Studies of lunar cryptomaria can also provide important information concerning the nature and origin of geochemical anomalies in orbital chemistry data sets [e.g., Hawke and Spudis, 1980; Hawke et al., 1985]. As discussed above, several mafic geochemical anomalies have been identified in the B-K region [e.g., Hawke and Spudis, 1980; Maxwell and Andre, 1981; Clark and Hawke, 1987]. It has been suggested that these mafic geochemical anomalies are due to the presence of cryptomare deposits in this area [e.g., Hawke et al., 1985; Antonenko et al., 1995]. Cryptomaria may prove to be responsible for mafic anomalies in the Lunar Prospector data sets elsewhere on the Moon [Hawke et al., 2003]. Finally, it has been suggested that cryptomare units played an important role in the formation of some lunar light plains deposits [e.g., Hawke et al., 1985; Head et al., 1993]. [10] Clementine multispectral images, Lunar Prospector (LP) gamma-ray spectrometer (GRS) data, and a variety of spacecraft imagery were used to investigate the composition and origin of geologic units in and around the Balmer- Kapteyn basin. The purposes of this study include the following: (1) to identify geochemical anomalies in the B-K region and to determine their origin, (2) to investigate the composition of surface units in the B-K basin region, (3) to identify and map the distribution of dark-haloed impact craters in the region, (4) to search for cryptomare deposits and to investigate the processes responsible for their formation, (5) to determine the compositions and ages of buried mare units, and (6) to investigate the processes responsible for the formation of the Balmer-Kapteyn light plains deposits. 2. Data and Methods [11] The primary data used for this investigation were images obtained by the Clementine UV-VIS camera. While the Clementine images provide high spatial resolution ( m/pixel) and show local and regional spatial relationships, they are of low spectral resolution. However, the band passes of the UV-VIS camera filters (415, 750, 900, 950, and 1000 nm) were chosen specifically for lunar study by the Clementine science team to maximize the information content returned by this multispectral instrument [Nozette et al., 1994]. [12] A Clementine five-color UV-VIS digital image model (DIM) for the Moon has been published on CD-ROM by the U.S. Geological Survey s Astrogeology Program [Isbell et al., 1999; Eliason et al., 1999, Robinson et al., 1999]. UV-VIS data from this DIM were mosaicked to produce Figure 3. Clementine 750 nm image mosaic showing the Balmer-Kapteyn region under high Sun illumination. A white box in Figure 1 outlines the area shown in this mosaic. The image mosaic is shown in simple cylindrical projection and has a spatial resolution of 1 km/pixel. Numbers indicate the locations of 25 dark-haloed impact craters (see Table 1). Black arrows indicate additional DHCs. HL1 and HL2 indicate the locations where spectra were obtained for highlands material (see Figure 6). MB1 shows the area where a spectrum was collected for a mare unit (Figure 6). image cubes centered on the Balmer-Kapteyn region with spatial resolutions of 1 km/pixel and 100 m/pixel (Figure 3). These image cubes, calibrated to reflectance, served as the basis for the production of a number of data products, including FeO and TiO 2 maps. [13] The FeO maps(figures 4b and 5b) were prepared using the algorithms of Lucey et al. [2000a]. The method developed by Lucey et al. [1995] for determining FeO abundances relies on 750 nm reflectance and 950 nm/750 nm ratio images to measure the spectral effects of ferrous iron in major lunar minerals such as olivine and pyroxene. This technique controls for the optical effects of the submicroscopic metallic iron that is produced as rocks are exposed to micrometeorite bombardment and solar wind implantation at the lunar surface [Lucey et al., 1995; Hapke, 2001]. The TiO 2 maps (Figures 4c and 5c) were produced using the method described by Lucey et al. [1998, 2000a]. This method utilizes a spectral parameter derived from 750 nm reflectance and 415 nm/750 nm ratio images. The mapping from the color-albedo parameter to wt.% TiO 2 is based on an understanding of the spectral effects of the Ti-rich opaque mineral ilmenite (FeTiO 3 ) as a component of a mineral mixture comprising the lunar regolith at the locations sampled by Apollo and Luna [Blewett et al., 1997; Lucey et al., 1998, 2000a; Jolliff, 3of16

4 Figure 4. (a) Optical maturity parameter image produced from Clementine UV-VIS images for the Balmer-Kapteyn region. This image, as well as those shown in Figures 4b and 4c, covers the area shown in Figure 3. The black vertical lines in this figure, as well as Figures 4b and 4c, are areas where data are missing. Brighter tones indicate lower maturity (fresher material). (b) FeO map of the B-K region. (c) TiO 2 map of the B-K region. 1999; Blewett and Hawke, 2001]. Optical maturity (OMAT) images (Figures 4a and 5d) were produced utilizing the algorithms of Lucey et al. [2000b] to provide information on the relative maturity of surface units in the B-K region. OMAT parameter data are useful for investigating the relative ages of deposits associated with dark-haloed impact craters. [14] Recent efforts have demonstrated that Clementine five-point spectra can be used to derive a number of diagnostic parameters that can be utilized to determine the lithology of areas for which the spectra were obtained [Tompkins and Pieters, 1999; Pieters et al., 2001]. We have used five-point spectra extracted from the registered and calibrated Clementine UV-VIS images (100 m/pixel) to determine the composition of geologic units in the B-K region (Figure 6). The methods described by Tompkins and Pieters [1999] were utilized. The dark-haloed impact crater (DHC) spectra shown in Figure 6 are the average of nine pixels selected to avoid steep slopes and to be representative of the dark ejecta. The DHCs for which these spectra were obtained are indicated in Figures 3 and 5a. In addition, spectra for two fresh highlands craters and one fresh mare crater in the B-K region are shown in Figure 6. The locations for which these spectra were obtained are indicated in Figure 3. 4of16

5 Figure 5. (a) Clementine 750 nm image (100 m/pixel) of the Phillips B crater area. A white box in Figure 8a indicates this area. Numbered arrows identify three well-developed DHCs (see Figure 3 and Table 1). Black arrows indicate other craters that expose mare basalts. (b) FeO map of the area shown in Figure 5a. (c) TiO 2 map of the area shown in Figure 5a. (d) Optical maturity parameter image produced for the Phillips B crater area. [15] The average FeO and TiO 2 values for the least contaminated portions of the dark halos are listed in Table 1 for the 25 DHCs shown in Figure 3. The values are the maximum values for the dark halos and they were obtained by averaging a 3 3 pixel (100 m/pixel) matrix in the area of the highest chemical concentration. These maximum values should most closely approximate those of the buried mare basalts. Care was taken to avoid steep slopes because of topographic effects that may affect the TiO 2 and FeO estimates. [16] Two Lunar Prospector orbital chemistry data sets were used to investigate the composition of surface units in the Balmer-Kapteyn region. The half-degree iron abundance data product contains data from the LP gammaray spectrometer acquired during the low-altitude portion of the mission [Lawrence et al., 2001, 2002a]. The absolute abundances are shown as FeO weight percent in Figure 7a. Themapbinsizeis0.5 by 0.5. A description of the reduction of this data set is given by Lawrence et al. [2001, 2002a]. It should be noted that LP-GRS FeO data were obtained for a greater regolith thickness than the FeO data derived from Clementine images. [17] The half-degree thorium abundance data set contains data from the LP-GRS obtained during the low-altitude portion of the LP mission. The Th abundances for the Balmer-Kapteyn region are shown in Figure 7b. The map 5of16

6 of these small DHCs (craters 4, 5 and 6) is provided in Figure 5a. The relatively small size of most of the DHCs may explain why so few DHCs were identified in previous studies of the B-K region [e.g., Schultz and Spudis, 1979; Hawke and Spudis, 1980; Maxwell and Andre, 1981]. [19] It is important to distinguish endogenic craters which may have dark halos of pyroclastic debris from exogenic impact craters. The criteria used in this study to positively identify dark-haloed craters of impact origin were derived from those given by Schultz and Spudis [1979] and Head and Wilson [1979]. These workers noted that exogenic craters are generally circular, have a depth/diameter ratio of about 1:5 for young craters <10 km in diameter, exhibit secondary craters, and have an uplifted rim. Endogenic craters often have a noncircular shape, are commonly aligned with fractures, rilles, or other lineaments, and do not exhibit raised rims, rays or secondary craters. Endogenic craters have depth/diameter ratios that are generally less than those of impact craters. These criteria were applied to the DHCs identified in the Balmer-Kapteyn region (Figure 3 and Table 1). No endogenic dark-haloed craters were Figure 6. Five-point Clementine UV-VIS spectra obtained for features in the Balmer-Kapteyn region. Representative spectra (HL1 and HL2, triangles) are shown for two fresh highland craters in the B-K region. A spectrum (MB1, diamonds) for a relatively young impact crater in a mare deposit is presented. Spectra (crosses) are presented for the dark ejecta of six well-developed DHCs (4, 5, 6, 10, 13, and 19) shown in Figure 3. bin size is 0.5 by 0.5, and the absolute abundances are given in units of ppm. The Th data presented in Figure 7b were produced by the ground-truth calibration method described in detail by Gillis et al. [2004]. This method uses an empirical correlation between Lunar Prospector gammaray data for Th and Apollo and Luna landing site soil compositions to yield absolute Th concentrations. This empirical ground-truth calibration for Lunar Prospector Th data provides self consistency between existing derived data and lunar-sample data. 3. Results and Discussion 3.1. Dark-Haloed Craters Identification, Distribution, and Origin of DHCs [18] Clementine 750 nm images and Apollo photographs were used to identify and map the distribution of darkhaloed craters (DHCs) in the Balmer-Kapteyn region. The Clementine 750 nm images are well suited for mapping DHCs because of their high spatial resolution and because they were obtained at a relatively low phase angle, which accentuates compositional contrasts. Large numbers of DHCs were identified in the B-K region. The locations of 25 of the best developed DHCs are indicated by numbers in Figure 3. The diameters of these craters are given in Table 1. Data for these 25 craters are given because their dark halos are prominent and complete. Additional DHCs are indicated by black arrows in Figure 3. While the DHCs listed in Table 1 range in diameter from 1 km to 39 km, most are less than 6 km in diameter. A high-resolution (100 m/pixel) view of several Table 1. FeO and TiO 2 Abundances for the Dark Halos of the Craters Shown in Figure 3 Crater Diameter, km Measured FeO, a wt.% Measured TiO 2, a wt.% Percent Mare Basalt b Calculated TiO 2, c wt.% ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± a These are the maximum values for the dark halos, and they were obtained by averaging a 3 3 pixel matrix in the area of the highest chemical concentration. b The percentage of mare basalt in the dark halo deposit was calculated using the method and assumptions given in the text (section 3.2.2). These percentages were calculated on the basis of the assumption that the buried mare basalts exposed by the DHCs contain 18.0 wt.% FeO. If the actual concentration of FeO in a given dark halo is less than the assumed value, the calculation will produce an underestimate of the percentage of mare basalt present in the deposit. c The concentration of TiO 2 was calculated for the mare basalts excavated by DHCs using the method described in the text (section 3.2.2). Since these TiO 2 values were calculated on the basis of the assumption that the buried mare basalts contain 18.0 wt.% FeO, they represent maximum values. 6of16

7 Figure 7. (a) This half-degree iron abundance map contains data from the LP-GRS acquired during the low-altitude portion of the mission [Lawrence et al., 2002a]. A mafic geochemical anomaly was identified in the Balmer region. LP-GRS FeO values for the B-K cryptomare range from 7% to 11%. Enhanced FeO values are also associated with cryptomare surfaces in the Lomonosov-Fleming (L-F), Schiller-Schickard (S-S), and Mendel-Rydberg (M-R) regions. The locations of these cryptomaria are indicated by white lines. (b) LP-GRS Th map for the B-K region. This map was produced using the method described by Gillis et al. [2004]. The map bin size is 0.5 by 0.5, and the absolute abundances are given in units of ppm. Humboldt, Petavius, Langrenus, and La Pérouse craters exhibit relatively low Th values. Most of the cryptomare surfaces in the B-K region have enhanced Th values. identified. It was determined that all of the dark-haloed craters were formed by impact. [20] The dark-haloed impact craters were further examined, as some dark halos surrounding impact craters are potentially deposits of low-albedo impact melt [e.g., Hawke and Head, 1977; Schultz and Spudis, 1979; Hawke et al., 1979]. Dark impact melt deposits are generally confined to narrow zones near the rim crests and are surrounded by high-albedo ejecta deposits. Schultz and Spudis [1979] noted that albedo contrasts between dark impact melts and bright ejecta deposits decrease rapidly with time, thus misidentification of impact melts as mafic dark halos should only be of concern for very young craters. Melts around very young craters will appear mature in optical maturity images [Hawke et al., 2002]. The dark-haloed impact craters identified in Figure 3 and listed in Table 1 do not exhibit proximal impact melt deposits. The dark halos were formed by the excavation of dark, mafic material from beneath a higher-albedo surface layer by the impacts. [21] Dark-haloed impact craters were identified in all sectors of the Balmer-Kapteyn region (Figure 3). However, the vast majority of the DHCs are located within the outer ring of the B-K basin (Figure 8b). The highest density of DHCs is associated with light plains deposits within the B-K inner ring and inside of Balmer and Phillips B craters (Figures 3, 5, and 8). The most prominent DHC in the 7of16

8 Figure 8. (a) Rectified Earth-based telescopic photograph of the Balmer-Kapteyn region (part of Plate 21-B of Whitaker et al. [1963]). The Phillips B crater area (Figure 5) is outlined by a white box. (b) Sketch map of the B-K region. The inner and outer rings of the Balmer-Kapteyn basin are indicated. The distribution of cryptomare is shown by a cross-hatched pattern. Major expanses of cryptomare are covered by the continuous ejecta of Langrenus, Petavius, Palitzsch B, Humboldt, and La Pérouse craters. These are indicated by A, B, C, D, and E. region (crater 10 in Figure 3) is 5 km in diameter and excavated dark material from beneath the continuous ejecta deposits of Petavius crater Composition, Maturity, and Spectra of DHCs [22] The FeO and TiO 2 maps (Figures 4b, 4c, 5b, and 5c) produced from Clementine UV-VIS images were used to determine the compositions of the DHCs in the B-K region. The average values for the least contaminated portions of the dark halos are listed in Table 1 for the 25 craters indicated in Figure 3. The FeO values range from 11.5 wt.% to 15.7 wt.%. These FeO values approach those exhibited by regoliths developed on typical nearside mare basalt flows (15 21 wt.%) [Lucey et al., 1998, 2000a]. It is important to bear in mind that the chemical values determined for mare surfaces by remote sensing techniques are for regoliths that may be contaminated by varying amounts of highlands debris [e.g., Korotev, 1998; Giguere et al., 2000, 2003]. The TiO 2 abundances of the dark halos listed in Table 1 range from 0.8 wt.% to 2.1 wt.%. These values are within the range of TiO 2 abundances determined for nearside mare surfaces by Giguere et al. [2000]. The FeO and TiO 2 values exhibited by the DHCs in the B-K region indicate that the dark halos are dominated by mare basalts. [23] The relative maturity of features in the B-K region is shown in Figures 4a and 5d. Bright features are less mature than dark features. The majority of dark halos associated with the DHCs are not bright in these OMAT parameter images (Figures 4a and 5d) indicating that the halos are fully mature. The low albedo of these DHCs is due to the composition of the darker, high-iron, mare material that has been excavated and become mature with age. Immature debris is, however, commonly exposed on the steep interior walls of these craters. A few DHCs have ejecta deposits that are less mature than the background material. These include 8of16

9 craters 4, 5, 6, 10, 13 and 14 (Figures 3, 4a, and 5d). The halos of these craters have not reached steady state maturity and will become even darker in time. [24] Five-point spectra were acquired for impact craters in the Balmer-Kapteyn region in order to investigate the mineralogy and lithology of a variety of geologic units. A comparison of DHC spectra with those obtained for highland and mare craters would allow us to confirm that mare basalts were excavated by dark-haloed impact craters and to assess the degree of contamination of DHCs by highland debris. An analysis of five-point spectra extracted for the ejecta deposits of several fresh impact craters in the highlands east and south of the Balmer-Kapteyn basin indicated that these craters expose material dominated by noritic anorthosite or anorthositic norite. Two representative highland spectra are shown in Figure 6 (HL1 and HL2). Both spectra have absorption bands centered near 0.90 mm, which suggests a mafic assemblage dominated by low-ca pyroxene. A spectrum for the ejecta of a relatively young impact crater in a mare unit is presented in Figure 6 (MB1). This spectrum exhibits a 1 mm absorption feature, and the band centered near 0.95 mm indicates a mafic assemblage dominated by high-ca clinopyroxene. [25] Spectra were obtained for six dark-haloed impact craters in the B-K region. They were acquired for the lowalbedo ejecta deposits surrounding the craters and are shown in Figure 6 (4, 5, 6, 10, 13, and 19). These spectra have 1 mm absorption bands that are also centered near 0.95 mm, indicating the dominance of high-ca pyroxene. Three spectra (4, 10, and 19) are very similar to the spectrum (MB1) collected for the fresh mare crater. The areas for which all four spectra (4, 10, 19, and MB1) were collected are dominated by relatively immature mare basalts. Three DHC spectra (5, 6, and 13) exhibit slightly higher reflectance values than those of the mare crater (MB1). The portions of the dark ejecta for which these spectra were obtained are composed of relatively immature mare debris contaminated with minor amounts of highland material. In summary, both the chemical data and the fivepoint spectra demonstrate that DHCs in the B-K region have excavated buried mare basalt from beneath a higher-albedo surface layer enriched in highland debris Cryptomare in the Balmer-Kapteyn Region [26] While the Balmer-Kapteyn region exhibits very few surface exposures of mare basalt, the large number of darkhaloed impact craters within the B-K basin (Figure 3) indicates the presence of laterally extensive mare deposits below the surface. A hidden mare deposit obscured by higher-albedo material is referred to as a cryptomare [Head and Wilson, 1992; Antonenko et al., 1995] Identification, Distribution, Thicknesses, and Ages of Cryptomare Deposits [27] Previous studies have proposed criteria for the identification of lunar cryptomaria [Schultz and Spudis, 1979; Hawke and Spudis, 1980; Hawke et al., 1985; Antonenko et al., 1995, Giguere et al., 2003]. A classification of evidence for cryptomaria identification was presented by Antonenko et al. [1995]. The most important criteria are (1) the presence of dark-haloed impact craters, (2) association with mafic geochemical anomalies, and (3) the presence of a significant component of mare basalt in the high-albedo surface unit as determined by spectral mixing analysis [e.g., Mustard et al., 1992; Head et al., 1993; Blewett et al., 1995]. [28] In order to determine the distribution of cryptomare deposits in the Balmer-Kapteyn region, we used the locations of dark-haloed impact craters (Figure 3) as well as FeO and TiO 2 maps (Figures 4b, 4c, 5b, and 5c). Other evidence for the presence of buried mare deposits included the identification of mare debris on crater walls and the occurrence of impact craters with incomplete or faint dark halos which exhibit enhanced FeO abundances. Mare debris on crater walls was identified on the basis of enhanced FeO values (see unnumbered arrows in Figure 5a). The distribution and extent of cryptomare in the Balmer-Kapteyn region is shown on the geologic sketch map presented as Figure 8b. It should be noted that strong evidence was required before a cryptomare deposit was mapped. Hence the cryptomare distribution shown in Figure 8b represents a fairly conservative estimate of the total extent of cryptomare in the B-K region. [29] Cryptomare deposits were identified in most sectors of the B-K region. Virtually all of the buried mare basalts are located within the outer ring of the Balmer-Kapteyn basin. The surface units within the inner basin ring are almost all underlain by cryptomare. Most of the cryptomare deposits in the B-K region are correlated with light plains units. These light plains deposits were divided into an older Imbrian- or Nectarian- age light plains (INp) unit and a younger, Imbrian-age light plains (Ip) unit by Wilhelms and El-Baz [1977]. Major expanses of cryptomare are also associated with the ejecta blankets of Petavius, Langrenus, Humboldt, La Pérouse, and Palitzsch B craters. These cryptomare deposits are indicated by the letters A, B, C, D, and E in Figure 8b. Minor amounts of cryptomare occur beneath other highland units in the B-K region. These other units include mantled and partly mantled terra material (units Nt and NpNt of Wilhelms and El-Baz [1977]). It should be noted that the mantled and partly mantled terra material mapped by Wilhelms and El-Baz [1977] does not exhibit a low albedo and should not be confused with dark mantle deposits of probable pyroclastic origin [e.g., Gaddis et al., 1985]. [30] The thickness of the basaltic fill in the B-K basin has long been the subject of controversy. Hartmann and Wood [1971] estimated a mean mare thickness of 2.5 km based on gravity data. A positive gravity anomaly in the B-K region was noted by Haines et al. [1978] and described in detail by Maxwell and Andre [1981]. Maxwell and Andre [1981] used this gravity data to argue that the mare fill was more than just a thin veneer as has been suggested by some previous workers. The dark-haloed impact craters in the region range in diameter from 1 km to 39 km (Table 1). With one exception, all are less than 6 km in diameter. Larger craters that have fully penetrated the mare layer generally do not exhibit a sharply defined proximal dark halo. A conservative estimate of the maximum depth of excavation of lunar impact craters is 0.1 of the diameter [Croft, 1980; Pike, 1977, 1980; Gillis, 1998]. Hence most mare basalts were excavated from depths of <0.6 km. However, the Nectarianaged crater Kapteyn B, which is 39 km in diameter, could have excavated mare material from depths as great as 3.9 km. Kapteyn B is located in an area near the center of the basin 9of16

10 (Figure 8) where previous workers [e.g., Hartmann and Wood, 1971; Maxwell and Andre, 1981] indicated that the mare fill was very thick. Although the thicknesses of the buried mare deposits appear to be <0.6 km in most areas, ancient mare basalt flows may be several kilometers thick in the central portion of the Balmer-Kapteyn basin. [31] It is important to determine the ages of the buried mare basalts in the B-K region. Large expanses of cryptomare are covered by an Imbrian-age light plains (Ip) unit. These obscured mare basalts must be of early Imbrian or greater age. Other cryptomare deposits are associated with an Imbrian- or Nectarian-age light plains (INp) unit. These buried basalts could have been emplaced no later than early Imbrian or late Nectarian time. Some cryptomare materials were excavated from beneath terra-mantling materials of Nectarian age (Nt). The mare basalts covered by the Nt unit must be of Nectarian or greater age. Kapteyn B (crater 19 in Figure 3) excavated mare basalt from beneath Nectarianaged and pre-nectarian-aged material of partly mantled terra (NpNt [Wilhelms and El-Baz, 1977]). Since mare debris was excavated from beneath a unit (NpNt) that is in part composed of pre-nectarian material, it is possible that some mare flows were emplaced in the Balmer-Kapteyn basin during the pre-nectarian period Compositions of Buried Mare Basalts [32] The compositions of the dark-haloed impact craters shown in Figure 3 were used to investigate the compositions of the buried mare units. The chemical data measured for DHCs in the Balmer-Kapteyn region (Table 1) must be used with caution because the dark halos may be contaminated by small but variable amounts of highland debris. This contamination may have occurred by impact mixing during the formation of the crater or by vertical mixing or lateral transport of highlands-rich material after the cratering event. Five-point spectra for six dark-haloed impact craters and other features in the B-K region are presented in Figure 6. Craters 4, 10, and 19 exhibit spectral parameters (e.g., relatively strong absorption bands centered near 0.95 mm) that are nearly identical to those derived for the spectrum (MB1) of a fresh mare crater. The dark halos of craters 4, 10, and 19 are composed of immature basaltic debris. In contrast, the spectra of craters 5, 6, and 13 exhibit higher reflectance values, and their spectral parameters indicate that the portions of the dark halos for which the spectra were obtained are composed of relatively fresh mare material contaminated by highland debris. The OMAT values exhibited by some of the other DHCs in the B-K region indicate that they are much older than craters 4, 5, 6, 10, 13, and 19. The dark halos of these older craters have been even more contaminated with highland material by vertical mixing and lateral transport from the surrounding terrain. [33] In an effort to make a first-order correction for the effects of highland contamination, we have listed only the maximum FeO and TiO 2 values for each dark-haloed crater in Table 1. These maximum values should most closely approximate those of the buried mare basalts. The FeO and TiO 2 values range between 11.5 wt.% and 15.7 wt.% FeO and from 0.8 wt.% to 2.1 wt.% TiO 2. It should be noted that 19 of the 25 DHCs listed in Table 1 have FeO abundances that are lower than those exhibited by regoliths developed on typical nearside mare basalts (15 21 wt.%) [Lucey et al., 1998, 2000a]. The DHC FeO values that best represent those of the buried mare deposits should be those obtained for the dark halos of craters that (1) have five-point spectra indicative of immature, uncontaminated mare material (craters 4, 10, and 19 in Figures 3 and 6), (2) have Earth-based telescopic spectra that indicate the presence of fresh mare basalt (crater 10 [Bell and Hawke, 1984]), or (3) exhibit OMAT values that suggest young deposits that should have experienced minimal contamination by vertical mixing or lateral transport from the surrounding highland terrain (e.g., craters 11 and 14). The DHCs that meet one or more of these criteria (craters 4, 10, 11, 14, and 19) exhibit FeO concentrations that range between 15.0 wt.% and 15.7 wt.%. Even if the alternate method of calculating FeO abundances from Clementine UV-VIS images recently presented by Gillis et al. [2004] is utilized, the FeO values for these DHCs do not exceed 18.0 wt.%. [34] In order to better estimate the actual TiO 2 concentrations in the buried mare basalts in the B-K region, additional efforts were made to correct for the effects of highland contamination of the dark halo deposits. We can estimate the concentration of TiO 2 in highland-free cryptomare material by assuming that the uncontaminated mare basalt averages 18.0 wt.% FeO and that the local highlands average 5.2 wt.% FeO. The latter value is based on measurements of the FeO values exhibited by the ejecta deposits of immature highland craters in the B- K region. The former value (18.0 wt.%) is higher than any of the FeO values measured for the dark halos in the region and is a reasonable estimate of the maximum amount of FeO present in the buried mare basalts. These FeO values allowed us to calculate the percentage of mare material in the ejecta surrounding each dark-haloed impact crater (Table 1). It should be noted that if the actual concentration of FeO in a given buried mare deposit is less than the assumed value of 18.0 wt.%, the calculation will produce an underestimate of the percentage of mare basalt present in the deposit. In turn, by assuming an average TiO 2 content in the highlands of 0.5 wt.%, and knowing the amount of mare material in the dark halo from the FeO calculation, we can estimate the TiO 2 concentration of a given cryptomare deposit. The calculated TiO 2 abundances for the DHCs in the B-K region are listed in Table 1. Since these TiO 2 values were calculated on the basis of the assumption that the buried mare basalts contain 18.0 wt.% FeO, they represent maximum values. If a given cryptomare deposit contains less than 18.0 wt.% FeO, its actual TiO 2 abundance would be less than the calculated value listed in Table 1. A histogram of the calculated TiO 2 values for the buried mare basalts that were exposed by DHCs in the B-K region is shown in Figure 9. These calculated TiO 2 values range from 1.0 wt.% to 2.5 wt.%. While a number of basalt classification schemes based on TiO 2 abundances are in current use [see Giguere et al., 2000; Warren, 2003], most schemes would classify the mare basalts exposed by DHCs in the B-K region as low-tio 2 basalts [Papike et al., 1976; Papike and Vaniman, 1978; Taylor et al., 1991; Neal and Taylor, 1992; Giguere et al., 2000]. The measured TiO 2 values obtained for three DHCs (Table 1) are <1.0 wt.%. When these values are corrected for the effects of highland contamination, the calculated TiO 2 values are >1.0 wt.%. Since these calcu- 10 of 16

11 Figure 9. Histogram of the calculated TiO 2 values for the buried mare basalts that are exposed by DHCs in the Balmer-Kapteyn region. lated TiO 2 abundances are maximum values, it seems likely that at least some very low TiO 2 (VLT) basalt was emplaced in the B-K region. [35] On the basis of Apollo X-ray fluorescence data and albedo information, Maxwell and Andre [1981] suggested that the basalts in the B-K region had compositions that were intermediate between average mare basalts and highland material. Hawke et al. [1985] presented a variety of remote sensing data which suggested that the buried mare material in the region had a composition that was similar to that determined for Apollo 17 sample [Ryder et al., 1977; Ryder and Spudis, 1980]. This high-alumina mare basalt is intermediate in composition between mare and KREEP basalts. [36] Aluminous (high-al) mare basalts are characterized by the highest alumina abundances of lunar mare basalts (Al 2 O 3 > 11 wt.%). High-alumina mare basalts were returned from the Apollo 12, 14, 17 and Luna 16 sites. A rich variety of aluminous mare basalts was collected from the Apollo 14 site. The Apollo 14 high-al basalts were returned predominantly as clasts in impact breccias although large samples and soil particles were identified in the Apollo 14 sample collection [Ryder and Spudis, 1980]. The old ages (>3.85 Ga) of the Apollo 14 aluminous basalt samples coupled with the fact that they commonly occur as clasts in breccias that were probably formed as a result of the Imbrium impact event indicate that these high-al mare basalts were emplaced on the lunar surface during the Nectarian and pre-nectarian periods [e.g., Hawke and Head, 1978; Ryder and Spudis, 1980]. The Luna 16 mission returned aluminous mare basalt fragments from Mare Fecunditatis. These basalts differ from the Apollo 14 high-al mare basalts in their higher TiO 2 abundances (up to 6 wt.%) and their younger (3.4 Ga) ages [Taylor et al., 1991; Kramer et al., 2004]. An Apollo 12 aluminous mare basalt sample (12038) has been dated at 3.1 Ga [Nyquist et al., 1981]. The existence of young high-al mare basalts in the Apollo 12 and Luna 16 sample collections suggests that aluminous mare volcanism spanned over a billion years [Kramer et al., 2004]. [37] At least some of the buried mare units in the Balmer- Kapteyn region may be high-alumina mare basalts. While we have no direct measurements of the Al 2 O 3 concentrations in these buried basalts, we do have estimates of their FeO and TiO 2 abundances derived from the analysis of dark-haloed impact craters described above (Table 1), and there is a well-developed anticorrelation between FeO and Al 2 O 3 in the lunar samples [e.g., Haskin and Warren, 1991]. High-alumina mare basalt samples generally have FeO abundances that range between 13 wt.% and 18 wt.% [Ryder and Spudis, 1980; Taylor et al., 1991; Kramer et al., 2004]. As discussed above, the fresh DHC FeO values that best represent those of buried mare deposits in the B-K region range between 15.0 wt.% and 15.7 wt.%. Most highalumina mare basalts exhibit TiO 2 concentrations less than 3.0 wt.%, but some samples have intermediate ( wt.%) TiO 2 values [Ryder and Spudis, 1980; Taylor et al., 1991; Giguere et al., 2000; Kramer et al., 2004]. The measured and calculated maximum TiO 2 values for DHCs in the B-K region range from 0.8 wt.% to 2.5 wt.%. These FeO and TiO 2 values are well within the range of values exhibited by aluminous mare basalts. [38] High-alumina mare basalts may be present in other lunar cryptomaria. Giguere et al. [2003] recently presented the results of a remote sensing study of dark-haloed impact craters and cryptomare deposits in the Lomonosov-Fleming (L-F) region of the Moon. The freshest DHCs within the L- F basin exhibit FeO values that range between 13.3 wt.% and 15.8 wt.%. The calculated TiO 2 abundances for all DHCs in the basin range between 0.5 wt.% and 4.2 wt.%. These FeO and TiO 2 abundances fall within the range of values commonly exhibited by aluminous mare basalts Compositions of Cryptomaria Surfaces [39] In order to better understand the processes responsible for the formation of cryptomare in the B-K region, LP- GRS FeO and Th abundance data (Figure 7) as well as TiO 2 and FeO maps produced from Clementine UV-VIS images (Figure 4) were used to investigate the surface compositions of the cryptomare deposits and other geologic units. The B- K cryptomare mapped in Figure 8b is outlined by a white line in the LP-GRS FeO map (Figure 7a). A mafic geochemical anomaly is associated with the Balmer-Kapteyn cryptomare. The LP-GRS FeO values for the cryptomare range between 7 wt.% and 11 wt.%. These FeO values are in good agreement with those determined using the Clementine FeO map (7 12 wt.%), and they are greater than the values exhibited by the surrounding highland terrain (Figures 4b and 7a). Clearly, major amounts of mare material are present in the surface of the B-K cryptomare unit. [40] An effort was made to estimate the amount of mare debris in the surfaces of the deposits that obscure the ancient basalt flows. Calculations were made using the techniques and assumptions described above for estimating the percentages of mare material in the ejecta surrounding dark-haloed craters. By assuming that pure, uncontaminated buried mare basalt averages 18 wt.% FeO and that the local highlands average 5.2 wt.% FeO, we can calculate the percentage of mare material in the regolith. The results indicate that the surfaces of the cryptomare deposits in the B-K region contain 14% to 53% mare debris. Most of the cryptomare surfaces contain 30% to 53% mare material. 11 of 16

12 [41] Since a mafic geochemical anomaly is associated with the B-K cryptomare, we used the LP-GRS FeO map to search for mafic anomalies associated with other lunar cryptomaria. Giguere et al. [2003] recently mapped the extent of cryptomare deposits in the Lomonosov-Fleming (L-F) region. The location of these buried mare deposits is shown by a white line (L-F) in Figure 7a. The cryptomare surfaces in the L-F region exhibit LP-GRS FeO values that range from 7 wt.% to 10 wt.%. These values are in good agreement with those measured using Clementine FeO images (7 11 wt.%) by Giguere et al. [2003]. There is a good correlation between the area of enhanced LP-GRS FeO abundances and the mapped cryptomare in the L-F region (Figure 7a). The enhanced FeO abundances exhibited by cryptomare surfaces in the L-F region indicate that major amounts of mare material were incorporated into the surfaces of the cryptomare deposits during the emplacement of the highlands debris that obscures the ancient mare basalts. Giguere et al. [2003] calculated that the surfaces of the cryptomare deposits in the Lomonosov- Fleming region contain 21% to 50% mare debris. They concluded that the high abundances of local mare basalt in the L-F cryptomare surfaces indicated that ballistic erosion and sedimentation played an important role in their formation. [42] A major expanse of cryptomare has been identified in the Schiller-Schickard (S-S) region of the Moon [e.g., Schultz and Spudis, 1979; Hawke and Bell, 1981; Mustard et al., 1992; Head et al., 1993; Blewett et al., 1995; Antonenko and Yingst, 2002]. The location of this cryptomare as mapped by Head et al. [1993] and Blewett et al. [1995] is outlined by a white line (S-S) in Figure 7a. A major mafic geochemical anomaly is associated with the S-S cryptomare. The LP-GRS FeO values for the S-S cryptomare surfaces range between 7 wt.% and 11 wt.%. These values are in general agreement with those measured from Clementine FeO images (7 13 wt.%) by Hawke et al. [2003]. Figure 7a shows that the LP-GRS FeO anomaly closely corresponds to the area of the S-S cryptomare as mapped by Head et al. [1993] and Blewett et al. [1995]. Spectral studies have demonstrated that the DHCs in the region excavated mare basalts from beneath light plains deposits emplaced as a result of the Orientale impact event [e.g., Hawke and Bell, 1981; Bell and Hawke, 1984]. Spectral mixing analyses have suggested that major amounts of mare material were incorporated into the light plains units by local mixing during the emplacement of Orientale basin ejecta in the Schiller-Schickard region [Mustard et al., 1992; Head et al., 1993; Blewett et al., 1995]. The LP-GRS FeO values (7 11 wt.%) exhibited by the cryptomare surfaces in the S-S region are consistent with a highlands-mare mixture. [43] Head et al. [1993] used Galileo SSI data to identify a cryptomare deposit in the Mendel-Rydberg (M-R) region (50 S, 95 W) which is located south of the Orientale basin on the west side of the Moon. Gillis [1998] used Clementine images to identify a much smaller area of buried mare basalt localized around the Copernican-age crater Guthnick (diameter = 36 km) in the M-R region. Gillis [1998] noted that Guthnick had excavated subsurface mafic material and concluded that this material was of intrusive igneous origin. The Mendel-Rydberg cryptomare as mapped by Head et al. [1993] is shown by a white line (M-R) in the Figure 7a. The LP-GRS FeO values for the region range from 7 wt.% to 9 wt.%. There is a good correlation between the area of enhanced LP-GRS FeO concentrations and the mapped cryptomare in the M-R region (Figure 7a). The iron anomaly is centered on the interior of the Mendel-Rydberg basin, a pre-orientale, multiringed impact structure. On the basis of the results of a spectral mixing analysis, Head et al. [1993] determined that cryptomare surfaces in the M-R region contained up to 40% mare debris. The enhanced LP-GRS FeO values provide independent support for the presence of major amounts of mare basalt in the regolith in the Mendel-Rydberg region. [44] In summary, major LP-GRS FeO enhancements are associated with cryptomaria in the Balmer-Kapteyn, Lomonosov-Fleming, Schiller-Schickard, and Mendel- Rydberg regions. These FeO anomalies correlate well with the mapped areas of the cryptomaria (Figure 7a). The enhanced LP-GRS FeO values associated with these extensive cryptomaria provide independent confirmation of the presence of significant amounts of mare debris in the regoliths developed on the cryptomare surfaces. Since four major cryptomaria exhibit anomalies in the LP-GRS FeO data set, the LP FeO map and other LP orbital geochemistry data sets may prove useful in identifying and characterizing smaller lunar cryptomaria. Finally, LP orbital geochemistry data for highland areas with cryptomaria should not be utilized in determinations of the average composition of the lunar highland crust. [45] Previous studies of Apollo and LP-GRS thorium data have identified Th anomalies in the Balmer-Kapteyn region. Haines et al. [1978] first reported an anomaly in this region. They identified an area with anomalously high Th concentrations (4.0 ppm) centered north of Balmer crater which correlated with Imbrian-age plains deposits. Clark and Hawke [1987] reported that Th concentrations greater than 2 ppm are exhibited by light plains in those portions of the B-K region for which Apollo orbital chemistry data were available. More recently, Lawrence et al. [2002b, 2002c, 2003] used LP-GRS Th data to identify and classify Th anomalies. They noted that unusually low Th abundances are associated with Langrenus and Humboldt craters. [46] A LP-GRS thorium map for the Balmer-Kapteyn region is presented as Figure 7b. Relatively low Th values are exhibited by Langrenus, Humboldt, Petavius, La Pérouse, and Palitzsch B craters. Low Th abundances are also exhibited by the cryptomare associated with the ejecta blankets of these craters (A, B, C, D, and E in Figure 8b). Higher Th values ( ppm) are exhibited by the remainder of the B-K cryptomare mapped in Figure 8b. The highest Th abundances ( ppm, white in Figure 7b) are associated with cryptomare deposits northwest of Kapteyn B crater. These Th values are similar to those reported by Clark and Hawke [1987] but are less than those (4.0 ppm) determined by Haines et al. [1978] for this area. In summary, most of the cryptomare surfaces in the B-K region exhibit enhanced Th abundances. This finding is consistent with previous suggestions that at least some of the buried mare basalts have thorium abundances that are higher than those typically associated with mare basalts in the lunar sample collection [e.g., Hawke and Spudis, 1980] Processes Responsible for the Formation of Cryptomaria [47] Four different processes have been identified that can lead to the burial or obscuration of a mare basalt deposit and the subsequent formation of a cryptomare [e.g., Schultz and 12 of 16