We used a Lunar Reconnaissance Orbiter Camera (LROC) global monochrome Wide-angle Camera (WAC) mosaic to conduct a survey of the Moon to search for previously unidentified pyroclastic deposits. Promising locations were examined in detail using LROC multispectral WAC mosaics, high-resolution LROC Narrow Angle Camera (NAC) images, and Clementine multispectral (ultraviolet-visible or UVVIS) data. Out of 47 potential deposits chosen for closer examination, 12 were selected as probable newly identified pyroclastic deposits. Potential pyroclastic deposits were generally found in settings similar to previously identified deposits, including areas within or near mare deposits adjacent to highlands, within floor-fractured craters, and along fissures in mare deposits. However, a significant new finding is the discovery of localized pyroclastic deposits within floor-fractured craters Anderson E and F on the lunar farside, isolated from other known similar deposits. Our search confirms that most major regional and localized low-albedo pyroclastic deposits have been identified on the Moon down to ∼100 m/pix resolution, and that additional newly identified deposits are likely to be either isolated small deposits or additional portions of discontinuous, patchy deposits.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Although volumetrically small compared to mare volcanism [e.g., Head, 1976; Head and Coffin, 1997], lunar pyroclastic volcanism nevertheless represents an important aspect of volcanic activity on the Moon [e.g., Wilson and Head, 1981, 1983]. A more complete understanding of the distribution, composition, and geologic setting of pyroclastic deposits across the Moon could provide important constraints on the timing, duration, and location of volcanic activity as well as insights into mechanisms of magma ascent and eruption and the composition and mineralogy of magma source regions. Identifying and characterizing the full range of compositions represented by pyroclastic deposits is an important part of efforts to understand the inventory of lunar surface materials. In addition, the unique composition and physical characteristics of lunar pyroclastic deposits make them valuable as potential raw materials for extraction of desirable resources such as iron, titanium, oxygen, and volatile elements [Hawke et al., 1990; Coombs et al., 1998].
 Previous investigators have placed lunar pyroclastic deposits into two general groups: “regional” and “localized” [Gaddis et al., 1985, 2003]. Regional deposits are typically larger, are found along the margins of maria, and are thought to have formed during Hawaiian-style fire-fountaining eruptions [Heiken et al., 1974; Wilson and Head, 1981; Gaddis et al., 1985; Weitz et al., 1998]. The regional deposits are commonly widely distributed and are often associated with a fissure or the source crater for a sinuous rille, and possibly formed during an earlier phase of lunar volcanic eruptions in which the maria were emplaced [Head, 1974]. Dating of pyroclastic material collected from regional deposits at Montes Apenninus (Apollo 15) and Taurus-Littrow (Apollo 17) indicated formation ages of ∼3.35–3.62 Ga [Spangler et al., 1984] and ∼3.48 Ga [Tera and Wasserburg, 1976], respectively. Localized deposits are smaller features interpreted to result from vulcanian-style eruptions caused by the explosive release of gases that accumulated under a solidified magma plug [Head and Wilson, 1979; Hawke and Head, 1980; Hawke et al., 1989]. Although there is no rigorous division in size between the two groups, regional deposits have surface areas that are typically greater than 1000 km2, while localized deposits are typically only a few tens to hundreds of km2 in size. Localized pyroclastic deposits have been identified in a variety of geologic settings. Many occur in floor-fractured craters, but others are found on or near major maria or adjacent to smaller mare deposits, and some are found as isolated deposits in the highlands [Gaddis et al., 1985; Hawke et al., 1989; Coombs and Hawke, 1992; Rosanova et al., 1998; Gaddis et al., 2000, 2003]. The ages of unsampled regional and localized pyroclastic deposits are difficult to assess, in part because their unconsolidated nature is thought to have reduced artificially the number of small craters observed on their surfaces [Lucchitta and Sanchez, 1975]. Stratigraphic evidence at sites such as the floor of Alphonsus crater suggests that these deposits post-date upper Imbrian-aged crater floor deposits [McCauley, 1969], but their exact ages are not at all clear.
 Soil samples collected during the Apollo missions contained glassy to microcrystalline beads, typically less than 0.1 mm in diameter, which have been interpreted to have formed in regional pyroclastic eruptions [e.g., Heiken et al., 1974; Pieters et al., 1974; Arndt et al., 1984]. The mineral compositions of the beads reflect a primitive, olivine-rich (picritic) composition, likely indicating both an origin deep within the lunar mantle (∼300–400 km) and rapid ascent and eruption [Delano, 1986; Shearer and Papike, 1993; Papike et al., 1998]. In addition, the surfaces of the beads are often coated or enriched with a variety of volatile elements (for example, S, In, Cd, Zn, Ga, Ge, Au, Pb, and Na) inferred to have been present in the eruption cloud [e.g., Baedeker et al., 1974; Meyer et al., 1975; Butler, 1978]. Measurements of the concentrations of volatiles trapped in the interior of a variety of volcanic glasses have revealed that H2O, F, S, and Cl are frequently present [Saal et al., 2008].
 Earth-based telescopic observations have identified a variety of spectral classes among regional and localized pyroclastic deposits. Regional deposits generally exhibit a broad ∼1-μm absorption feature with a shape and position that are consistent with a mixture of pyroxene with olivine and/or (based on their low albedos) iron-bearing volcanic glasses [e.g., Gaddis et al., 1985]. Variations in the strength of the ∼1-μm feature may be related in part to the degree of crystallinity of the pyroclastic beads [Weitz et al., 1998]. Localized pyroclastic deposits have been classified into three groups based on the steepness of the spectral continuum slope and the position, depth, and shape of the ∼1-μm band [Hawke et al., 1989; Coombs and Hawke, 1992]. These spectral groups were interpreted to represent varying contributions of highlands wall rock, basaltic plug rock, and juvenile basaltic magma to the mixture of clasts emplaced at each location.
 Pyroclastic deposits have also been examined using Earth-based and satellite-based radar [e.g., Zisk et al., 1974; Gaddis et al., 1985; Campbell et al., 2008; Carter et al., 2009, 2010; Trang et al., 2010]. Radar provides an alternative approach for studying the properties of the lunar surface that is complementary to spectroscopic methods. The radar signal is sensitive to both the composition of the surface and the presence of surface or subsurface scatterers on the scale of the radar wavelength. The ability to detect such scatterers at the centimeter to meter scale has direct applicability to identifying and characterizing smooth textured, relatively block free pyroclastic deposits. The coverage and spatial resolution of the radar data sets produced to date has limited their application to studies of relatively large and thick pyroclastic deposits. Consequently, radar data were not employed in this study, which focused on identifying smaller and typically thinner pyroclastic deposits. However, as more complete radar coverage is acquired at higher resolutions, it may provide a valuable tool for characterizing smaller pyroclastic deposits as well.
 Recent lunar missions have resulted in a vast increase in both the quantity and quality of remote sensing data. Increased image resolution and coverage have enabled a more thorough and detailed search for potential pyroclastic deposits. For this study, we used Lunar Reconnaissance Orbiter Camera (LROC) data to search for potential localized pyroclastic deposits that have not been previously identified. Our analysis takes advantage of the high resolution, global coverage provided by LROC as well as our increased ability to compare multiple data sets as each new mission is flown to the Moon. The goals of this effort were to assess the utility of the LROC data in locating and studying pyroclastic deposits; to estimate the prevalence of uncatalogued low-albedo pyroclastic deposits on the Moon; to help compile a more complete catalog of pyroclastic deposits; and to characterize the properties and possible origins of these features.
 The Lunar Reconnaissance Orbiter (LRO) Wide-angle Camera (WAC) acquires multispectral images at two ultraviolet (UV) and five visible (VIS) wavelengths (320, 360, 415, 565, 605, 645, and 690 nm) at a resolution of ∼400 m/pix in the UV and ∼75 m/pix in the visible during the mission's primary mapping orbit phase. The LRO Narrow Angle Camera (NAC) produces monochrome images at a resolution of ∼0.5 m/pix [Robinson et al., 2010] during the same mission periods. We examined a preliminary 100 m/pix global monochrome (645 nm) WAC mosaic of the Moon [Robinson et al., 2011; Denevi et al., 2011; Sato et al., 2011; Speyerer et al., 2011] for dark-toned deposits with morphologic indicators of pyroclastic origin. Six key criteria employed in the identification were: (1) Material exhibits smooth texture, and is block-free in high-resolution images; (2) material mantles and subdues subjacent terrain; (3) deposit exhibits diffuse margins; (4) material does not flood adjacent topographic lows; (5) deposit surface does not exhibit lava flow textures; and (6) deposit occurs in association with rilles, fractures, or other possible vents [e.g., Gaddis et al., 1985; Hawke et al., 1989].
 Our search focused initially on locations in the LROC targeting database [e.g., Gaddis et al., 2009] which had characteristics consistent with potential pyroclastic deposits. This database was open for target entry by a wide range of individuals based on both published and unpublished target characteristics. Key information often used to identify possible pyroclastic deposits included prior mapping, unusual albedo, and proximity to fractures, rilles, or possible vents. These deposits typically were not well enough resolved in previous data sets to assess their mode of emplacement. For the purpose of this survey, we avoided well-studied locations that have previously been identified as pyroclastic deposits. In addition to this targeted search, we also examined the new global WAC mosaic using a gridded search between 60°N and 60°S latitude; higher latitude locations were excluded due to unfavorable illumination geometry. For locations that met our preliminary morphologic criteria, we processed and examined WAC color mosaics [Robinson et al., 2010] and/or high-resolution NAC images (if available for the target area) to look at additional details of mantling relationships, deposit textures, rock abundances, and the possible presence of volcanic vents.
 We also examined Clementine spectral reflectance (CSR) data at ultraviolet, visible and near-infrared wavelengths (415, 750, 900, 950, 1000 nm) for candidate pyroclastic deposits (http://www.lpi.usra.edu/lunar/tools/clementine/instructions/UVVIS_DIM_Info.html). The position of the filters in the spectrum was chosen to maximize information about absorption features related to common lunar minerals such as plagioclase feldspar, orthopyroxene, clinopyroxene, and olivine [e.g., Tompkins and Pieters, 1999]. We reviewed Clementine UVVIS “natural” color composite images (red = 1000 nm, green = 900 nm, blue = 415 nm) to assist in delineating the extent of dark-toned material. We used the 750 nm reflectance data to estimate the Lambert albedo to compare the albedo of candidate pyroclastic deposits with that of surrounding deposits. We also used the UVVIS color ratio composite images and derived iron abundance images available at ∼100 m/pixel from the PDS/USGS Map-A-Planet database (http://www.mapaplanet.org) [e.g., Garcia et al., 2010]. These data provide compositional information that can help delineate the extent of dark mantle deposits by differentiating them on the basis of iron content from subjacent materials, especially if the mantling material is thin or shading effects obscure albedo differences. The ∼1 μm absorption band in CSR data results primarily from Fe2+ present in components such as orthopyroxene, clinopyroxene, olivine, or iron-bearing volcanic glass [Pieters and Englert, 1993]. Although the spectral resolution of the CSR data does not allow for differentiation among these possible components, the band depth at 950 nm has been shown to be correlated to the total abundance of these mafic components [e.g., Lucey et al., 2000]. The CSR color ratio composite images are created by displaying the 750/415 nm ratio as red, the 750/950 nm ratio as green, and the 415/750 nm ratio as blue. Generally, spectrally mature highland deposits appear red, and “fresh” (spectrally immature) highlands exposures appear blue. Basaltic mare-type deposits appear yellow to orange (iron-rich, lower-Ti) or blue (iron-rich, higher-Ti) [McEwen et al., 1994; Pieters et al., 1994]. It is important to note, however, that individual colors do not correspond to unique compositions, and regions containing heterogeneous compositions and maturity states can appear complex in the ratio images. The derived iron abundance images are created using the algorithms of Lucey et al.  as modified by Lawrence et al. . Weight percent FeO is presented using a gray scale ranging from 0 to 25 wt.%.
 The morphology and sharpness of deposit margins provide important clues to their mode of emplacement. Effusive lavas often pool in topographic lows, resulting in smooth, level deposits with relatively sharp margins that follow the contours of the preexisting topography [e.g., Wilhelms, 1987]. This can be seen when a younger deposit embays older, higher terrain within depressions or valleys, for example. In contrast, pyroclastic fall deposits mantle the preexisting topography, covering both high and low areas [Wilson and Houghton, 2000]. Fall deposits can also produce diffuse margins as the thickness of material emplaced tapers off with increasing distance from the source vent [e.g., Wilson and Head, 1981]. The bedrock block abundance at the surface of a deposit is also useful for distinguishing lava flows from pyroclastic deposits. For example, fine-grained pyroclastic deposits will likely contain blocks only very close to the source vent or where impacts have excavated blocky bedrock from beneath the pyroclastic layer. In a similar manner, the presence of flow textures such as ripples, ropy surfaces, and lobes provides evidence of emplacement as either impact melt or an effusive, rather than pyroclastic, deposit [Schaber, 1973; Wilhelms, 1987]. Care must be taken when evaluating these factors, however. Fall deposits shed from slopes can follow topographic contours [Wilhelms, 1987], and impact gardening can “smear” the edges of effusive deposits, complicating efforts to differentiate between processes. In addition, ballistically emplaced impact melt can “splash” over uneven terrain, resulting in an appearance similar to pyroclastic mantling [e.g., Cintala and Grieve, 1998].
 Volcanic vent characteristics can provide important evidence regarding the mode of eruption. For effusive eruptions, source vents are often buried by the late stage lavas [Greeley, 1976; Wilhelms, 1987]. When they are visible, they typically are associated with rilles or domes. In contrast, while pyroclastic source vents can also be associated with these features, they are more often observed as irregular craters or depressions similar to maars on Earth, often in association with fractures or graben [Head and Wilson, 1979]. There is some overlap in the morphologies of both kinds of eruptive styles, however, and both types of volcanic source vents/regions can be obscured by subsequent eruptions or by impact ejecta.
 After performing an initial screening of candidate pyroclastic deposits using the data sets and criteria described above, we eliminated many from further consideration because they lacked sufficient evidence of pyroclastic emplacement (classified as “unlikely”) or the available data did not support a conclusive interpretation (classified as “inconclusive”). Other potential mechanisms for formation of irregular or patchy dark deposits include partially obscured mare material or partially exhumed cryptomaria [e.g., Antonenko et al., 1995], small mare ponds, and impact melt [e.g., Cintala and Grieve, 1998]. Many of the locations classified as inconclusive had high incidence angle illumination that was not optimal for assessment of albedo features. The remaining locations are considered to be probable pyroclastic deposits based on a qualitative assessment of the six criteria listed above. These are presented as examples and not as a comprehensive list of remaining pyroclastic deposits.
3. Results and Discussion
 Preliminary review of approximately 125 dark-toned deposits resulted in a list of 47 locations that warranted closer inspection. Of these 47 deposits, 12 were assessed as probable newly identified pyroclastic deposits (Table 1) and 35 were assessed as inconclusive or unlikely (Table 2). Probable or possible pyroclastic deposits are grouped geographically as follows: the farside Freundlich-Sharonov basin, Orientale basin, Mare Nubium region, and the Montes Carpatus region (Figure 1). The global distribution of these potential pyroclastic deposits is presented in Figure 1 along with the locations of many previously identified pyroclastic deposits as compiled by Gaddis et al. . Previously identified pyroclastic deposits shown on more detailed figures throughout the paper are identified by letters and listed in Table 3.
Table 1. Summary of Newly Identified Lunar Pyroclastic Deposits
 Anderson E and F are Nectarian craters with diameters of 28 km and 49 km, respectively, located in the highlands on the farside east of Mare Moscoviense (Figures 1 and 2). They lie near the center of the 600 km diameter Freundlich-Sharonov basin (Figure 2). The unusual Buys-Ballot feature, discussed below, is located approximately 100 km to the north. Anderson E and F both exhibit a series of floor fractures aligned primarily concentric to the crater rims, and both crater floors appear raised with respect to the typical simple crater profile. Numerous small dark deposits are revealed in the WAC image mosaic (Figure 3a), with many of the dark spots associated with craters that appear to be endogenic based on several of the criteria established by Head and Wilson : their irregular shape, alignment along floor fractures, lack of rays, and lack of a raised rim. Some of these deposits may have been obscured by the emplacement of highlands material in the form of crater rays crossing Anderson E and F. The dark material surrounding these possible vents has mantled the hummocky floor material of both Anderson E and F. Typical Lambert albedo of these deposits ranges from 0.201 to 0.213, compared to 0.262–0.267 for the surrounding crater floor deposits (Table 4). The enhanced mafic signature of the dark mantling material can be seen in the CSR ratio image (Figure 3b). Green colors seen across the center of Anderson E and at the spots in the southern portion of Anderson F indicate mixing of highland debris with mafic materials. Approximately 10 possible vents have been identified in Anderson E, to the north, and at least four possible vents have been identified in Anderson F. Examples of possible volcanic source fissures are indicated on the NAC image in Figure 4. We have concluded that these are likely pyroclastic deposits based on the dark tone, association with possible vents located along floor fractures, areal distribution around the possible vents, and their mafic signature.
 The raised floors and the system of floor fractures within Anderson E and F suggest post-impact intrusion of lava beneath the crater floor [e.g., Schultz, 1976; Stewart-Alexander, 1978]. The dark-mantle deposits in Anderson E and F appear to be Alphonsus-type deposits erupted from a series of small vents aligned with the fracture system, suggesting their origin in volatile-rich vulcanian-style eruptions containing relatively small amounts of juvenile material [e.g., Head and Wilson, 1979]. The two small deposits within the southern portion of Anderson F may represent gas-rich eruptions that emplaced very little juvenile material; the green colors in the CSR ratio image suggest that these deposits are a mixture of mafic minerals and highlands material that could represent wall rock eroded from the volcanic vents [e.g., Hawke et al., 1989]. The presence of such volcanic features within the central farside highlands indicates that magma ascent and eruption have occurred even within this region of relatively thick crust [e.g., Kaula et al., 1974; Zuber et al., 1994], likely due to localized crustal thinning (∼10 km thickness) related to the formation of the Freundlich-Sharonov basin [Wieczorek et al., 2006]. However, this is the only such occurrence that we have located, and it appears to represent an endpoint in the continuum of eruption styles, where the eruption was just barely able to reach the surface but could not transport enough magma to the surface to form an effusive deposit [e.g., Head and Wilson, 1979].
 Buys-Ballot is an unusual oblong Imbrian impact crater consisting of a roughly circular northern part (D = 55 km) and a narrower lobe extending an additional 20 km to the south, giving it a pear shape [Schultz et al., 1998; van der Bogert et al., 1998]. It has a central ridge running along the long axis and contains a mare-like flat floor deposit in the northern portion. Immediately south and southeast of Buys-Ballot is a dark deposit (Lacus Luxuriae) that consists of a very dark-toned central region surrounded by lobes of lighter-toned deposits to the west, south, and east (Figure 5a). The southern portion of Lacus Luxuriae appears to be relatively smooth-surfaced and constrained primarily by local topography; we interpret this region as effusively emplaced mare basalt. In contrast, the dark-toned material to the north, where we identify a candidate pyroclastic deposit, appears to mantle a portion of the highlands north of Lacus Luxuriae as well as hummocky terrain south of Buys-Ballot and a portion of the southern wall of the crater itself (Figure 5b). The dark tone of this mantling material can be seen in the CSR UVVIS image (Figure 6a); a typical Lambert albedo within the deposit is 0.181, compared to 0.245 for the surrounding highlands deposits (Table 4). The derived iron image (Figure 6b) shows that this area contains relatively iron-rich material similar in composition to the smoother deposits within the southern and eastern portions of Lacus Luxuriae. We have identified this as a likely pyroclastic deposit based on the observed mantling of the crater wall, hummocky terrain, and adjacent highlands by the dark-toned, iron-rich material.
 Buys-Ballot most likely formed from an oblique impact coming from the north, and the dark-toned deposit south of the crater (Lacus Luxuriae) has been interpreted as residual material from a mafic-rich impactor deposited along the downrange rim of the impact crater [Schultz et al., 1998]. These authors cited the region's remoteness from known volcanic activity in support of a non-volcanic interpretation of this feature. The relative paucity of volcanic deposits over much of the lunar farside is generally attributed at least in part to greater crustal thickness and higher surface elevation relative to the nearside, which would inhibit the ascent of magma to near-surface depths [e.g., Head and Wilson, 1979, 1992; Zuber et al., 1994]. However, farside volcanic deposits are found in basins (e.g., South Pole–Aitken, Moscoviense, Orientale) where excavation and crustal thinning have occurred as a result of the basin-forming impacts. Although the Freundlich-Sharonov basin is not as deep as the mare-containing farside basins [Scholten et al., 2011], the evidence for pyroclastic eruptions at Anderson E and F (section 3.1.1) supports the possibility of volcanism at Buys-Ballot as well. This would suggest that the depth of basin excavation was sufficient to allow magma ascent to near-surface magma chambers where subsequent eruption could occur [e.g., Head and Wilson, 1979]. We believe this is a plausible alternative to the Schultz et al.  hypothesis that the mafic deposits within Buys-Ballot and Lacus Luxuriae were deposited by the crater-forming impactor.
 Kopff is a 41-km Imbrian-aged impact crater located in the eastern portion of the Orientale basin, immediately east of Mare Orientale but within the Montes Rook ring (Figures 1 and 7) [Scott et al., 1977; Whitten et al., 2011]. Scott et al.  mapped mare materials in the floor of Kopff as being Eratosthenian/Imbrian in age, indicating that they formed at the same general time as the mare deposits in Mare Orientale to the west and Lacus Veris to the north and east. A number of pyroclastic deposits have been previously identified in the vicinity of Orientale basin; besides the well-known Orientale dark ring deposit to the southwest [e.g., Head et al., 2002], there are several pyroclastic deposits east and northeast of Kopff (e.g., Schluter A, Hedin, and Riccioli) [Coombs and Hawke, 1992]. Possible pyroclastic deposits at Schluter, approximately 300 km northeast of Kopff, are discussed below.
 There are a series of rilles (Rimae Kopff) in the floor of the crater, as well as several irregular craters and depressions concentrated in the southern portion of the crater floor (Figure 7a). A dark-toned deposit visible on the southern floor of Kopff appears to extend onto the rim of the crater, suggesting a pyroclastic component to this deposit. A typical Lambert albedo of the mantling material on the rim is 0.146, compared to 0.217 for the surrounding highlands deposits (Table 4). Inspection of the Clementine derived iron and color ratio composite images (Figures 7c and 7d) indicates a compositional affinity between the dark-toned floor material and the dark-toned material along the southern rim, which is more mafic (orange) in the ratio image and higher in iron than adjacent rim deposits. Although the iron abundance is highest in the portions of the crater floor that appear to contain effusive deposits, virtually the entire floor and much of the crater wall exhibits elevated iron compared to the surrounding highlands (Figure 7c). In a similar manner, the CSR color ratio image exhibits green and orange hues across much of the crater floor and walls, indicating widely distributed mafic material that is not confined to the regions of smooth effusive deposits (Figure 7d). NAC mosaics of the southern floor and rim indicate that dark material mantles the ridges and hummocky terrain (Figures 8 and 9). Partially buried large boulders indicate the presence of a thick layer of fine-textured regolith (Figure 9). NAC images of the crater floor show deep, irregular pits within several rilles (Figures 8 and 10) which are possible source vents for the pyroclastic material (e.g., Figure 10b). We have identified this as a likely pyroclastic deposit based on 1) the observed mantling of the hummocky terrain on the crater floor and the southern crater wall and rim by dark-toned, mafic, iron-rich material, and 2) the association of these deposits with possible volcanic vents aligned with fractures on the crater floor.
 Kopff crater has a somewhat unusual morphology (shallow floor, no central peak, subdued rim, lack of ejecta rays) that has led some to suggest that it may have been formed by impact into partially molten Orientale basin materials [e.g., Spudis et al., 1984]. The raised and fractured floor materials indicate subsequent modification by volcanic and/or tectonic processes; however, the unusual conditions of its formation make it difficult to differentiate impact melt materials from volcanic materials. The occurrence of a pyroclastic deposit in such a setting may be unique.
 Schluter is an 89-km Imbrian-aged impact crater located northeast of Orientale basin along the western limb of the Moon (Figure 1). It lies within the continuous Orientale ejecta blanket just outside the Montes Cordillera ring [Scott et al., 1977]. The crater has a central peak surrounded by hummocky floor material, which has been flooded by mare flows in the northern half of the crater (Figure 11). Scott et al.  mapped the mare materials as being Eratosthenian/Imbrian in age, indicating they formed at the same general time as the mare deposits in Mare Orientale to the west and Lacus Autumni to the southwest. There is a network of rilles on the crater floor, some of which appear to be partially inundated by mare materials. A number of pyroclastic deposits have been previously identified in this region of the Moon [e.g., Coombs and Hawke, 1992; Gaddis et al., 2003]. Coombs and Hawke  mapped pyroclastic deposits within Schluter A and at Mare Autumni, both located within 75 km south/southeast of Schluter. They also mapped pyroclastic deposits associated with mare fill deposits within the craters Hedin and Riccioli northeast of Schluter.
 We have identified three areas of additional possible pyroclastic deposits within Schluter: two associated with the mare materials in the eastern and northeastern portions of the crater floor (#5b, #5c) and a third possible deposit just west of the central peak (#5a) (Figure 11). Deposit #5a appears as dark mantling surrounding a depression and fracture that could be a source vent. The primary diagnostic feature of these pyroclastic deposits is their mantling relationship to the hummocky floor materials. A typical Lambert albedo of the mantling material is 0.143, compared to 0.197 for the surrounding light-toned crater floor deposits (Table 4). The mantling material appears to be similar in composition to the surrounding mare fill, as shown by the CSR ratio and derived iron images (Figures 12a and 12b).
 Schluter is one of a number of highland craters containing basalt flows with possible pyroclastic mantling (e.g., Petavius, Humboldt). This could be a relatively common occurrence on the Moon, but identification of mantling on dark-toned mare deposits is often difficult due to the lack of a strong albedo contrast and the generally smooth topography. Mantling of the effusive deposits by pyroclastic materials indicates that the explosive volcanism came at the end of the period of volcanic activity, perhaps due to stalling and degassing of the final magma pulse beneath the surface.
 Mare Nubium, on the south-central nearside, exhibits little evidence of the pyroclastic activity that is associated with other nearside maria such as Serenitatis, Imbrium, Vaporum and Humorum. The most notable pyroclastic features in this region are the dark-halo craters of Alphonsus, the type locality for localized pyroclastic deposits [Head and Wilson, 1979]. Previously recognized deposits also include localized deposits at Fra Mauro, Lassell crater, Airy, and Hell [Weitz and Head, 1999; Gaddis et al., 2003]. We identified additional pyroclastic deposits in the maria at Birt E and in the highlands southeast of Nubium at Walther A (Figure 13).
3.3.1. Birt E
 Birt E (Figure 14) is a small deposit associated with a linear rille. Possible pyroclastic mantling was identified during geologic mapping of this region [Holt, 1974]. The Birt E DMD surrounds the crater Birt E (D = 5 km), which is the largest of several pits located along Rima Birt I. The concentric nature of the deposit suggests that this is the likely source for at least some of the dark mantle material. However, the morphology of the dark-toned deposit does not rule out the possibility that additional material may have erupted from the second linear rille located immediately west of Birt E. A typical Lambert albedo of the dark-toned deposit is 0.103, compared to 0.123 for the surrounding maria (Table 4). In the Clementine color ratio composite image (Figure 14b), this deposit appears orange, suggesting a lower titanium concentration than the surrounding bluer mare deposits. We have identified this as a likely pyroclastic deposit based on the dark tone, the association with a likely vent, and the contrast in composition with the surrounding maria.
 The Birt E crater and nearby rilles are similar to features found in association with many mare deposits. These have been interpreted as volcanic vents and lava channels related to the effusive eruptions which emplaced the maria [e.g., Head, 1974]. It is likely that early stage vents and rilles were buried by subsequent eruptions; therefore features such as Birt E and Rima Birt I must have been products of late stage volcanic activity. Similar to the deposits in western Montes Carpatus described below, the Birt E pyroclastic deposit may be a late-stage volcanic deposit formed by the release of volatiles during cooling of residual magma after emplacement of the regional mare basalt flows. In this model, the pyroclastic material would contain a large component of basaltic plug rock fragmented and erupted by the accumulated gas pressure [e.g., Hawke et al., 1989].
3.3.2. Walther A
 Walther A is an 11-km Imbrian-aged impact crater located in the south-central highlands southeast of Mare Nubium (Figure 1). It was formed in the hummocky pre-Imbrian floor of the larger (D = 135 km) crater Walther (Figure 15a) [Pohn, 1972]. To the south and west of Walther A, the floor of Walther consists of a smooth plains unit. The floor of Walther A itself is slightly hummocky and does not exhibit evidence of significant, if any, resurfacing by mare basalt. There is a thin, diffuse dark mantling deposit visible in the WAC image (Figures 15b and 15c) across most of the floor, walls, and rim of the crater, and extending in east and northeast-trending lobes for 1–2 crater diameters. A typical Lambert albedo of the mantling material is 0.202, compared to 0.263 for the nearby Walther crater floor deposits (Table 4). The extent of this material is seen more clearly on the Clementine color ratio composite and derived iron abundance images (Figures 16a and 16b), due to its relatively mafic and iron-rich composition. The Walther A DMD is somewhat unusual among potential pyroclastic deposits in that it occurs in the highlands, away from major effusive deposits, with no significant associated crater floor fractures and no obvious candidate source vents.
 We favor an interpretation of the origin of the dark-toned material at Walther A as pyroclastic, rather than impact melt or excavated cryptomaria, for several reasons. The asymmetric distribution of the dark-toned material is not consistent with an origin as ejecta from Walther A. The DMD also mantles the topography formed by younger impacts into the Walther A ejecta blanket, seen most clearly in the two 3–4 km craters immediately north and northeast of Walther A (arrows, Figure 15b), indicating that the dark deposit is significantly younger than Walther A itself. Impact melt tends to form irregular, patchy deposits in topographic lows rather than continuous lobes, and would be similar in composition to the surrounding terrain [Hawke and Head, 1977; Bray et al., 2010]. Impact melt deposits also degrade relatively quickly, and are therefore not typically seen in association with Imbrian-aged craters such as Walther A. The darkest portion of the DMD is centered near the north rim of Walther A (Figure 15c), and a possible source of the pyroclastic material is within the irregular craters and depressions found there. We have identified this as a likely pyroclastic deposit based on the observed mantling of the hummocky terrain on the crater floor, the northern crater wall and rim, and the terrain northeast of Walther A by dark-toned, mafic, iron-rich material.
3.4. Montes Carpatus
 The Montes Carpatus are located along the southern border of Mare Imbrium, immediately north of Copernicus crater (Figure 17). They are part of the raised rim structure of the Imbrium basin, and have experienced significant resurfacing by Copernican ejecta [Schmitt et al., 1967]. Extensive regional pyroclastic deposits are located east and southeast of Montes Carpatus at Sinus Aestuum and Rima Bode [e.g., Gaddis et al., 2003]. A localized pyroclastic deposit (Figure 17, locations C1–C4) has previously been identified within Montes Carpatus [Gaddis et al., 1985], one of several locations in the region mapped by Schmitt et al.  as volcanic materials of possible pyroclastic origin.
 The dark mantle deposits in the vicinity of Montes Carpatus occur as irregular patches, either within the highlands or on the maria adjacent to the highlands (Figure 17). The irregular shape of these deposits is in part due to the presence of bright rays (apparently superimposed) of highland debris from Copernicus, which appear to partially obscure larger areas of dark material. Deposits near Tobias Mayer (#8–10) occur within generally flat effusive deposits and mantle hills that appear to be remnant blocks of highland material (Figures 18 and 19). Deposit #8 mantles blocks of highland material that were previously identified as a mare dome [Schaber, 1973]. However, close inspection of low-sun Apollo images indicates that what might appear to be a dome with a central pit is actually two separate highland blocks (Figure 18b). Deposit #9 mantles numerous small hills that lack summit craters and appear to also be remnants of highland material (Figure 18c). Deposit #10 mantles a mixture of small hills and hummocky terrain (Figure 19c). Typical Lambert albedo of these deposits ranges from 0.103 to 0.106, compared to 0.125–0.130 for the surrounding light-toned streaks (Table 4). All these deposits exhibit an enhanced mafic signature in the CSR ratio images (Figure 19), although the contrast in composition with the surrounding mare materials is limited. We identified these locations near Tobias Mayer as likely pyroclastic deposits based primarily on the dark tone and the observed mantling of the blocks of highland material preserved among the mare deposits.
 Deposits near Gay-Lussac (#11–12) occur within the mountains immediately north of Copernicus (Figures 17 and 20). The dark material is somewhat patchy with mostly diffuse margins, but occasionally more defined edges are seen (Figure 21a). These deposits occur in hilly terrain with steep slopes, many of which exhibit layers and streaks of dark material. In places, this material also appears to have been obscured by or mixed with Copernicus ejecta. Typical Lambert albedo of these deposits ranges from 0.109 to 0.122, compared to 0.154–0.155 for the surrounding light-toned streaks (Table 4). CSR ratio images of the Gay-Lussac deposits reveal that the dark mantle material has an enhanced mafic signature compared to the surrounding unmantled highlands (Figure 20b), although the deposits have a mottled appearance in the ratio images due to the complex character of the terrain. We identified these locations near Gay-Lussac as likely pyroclastic deposits based on the observed mantling of the hilly terrain by dark-toned material, the presence of very dark streaks on many of the steep slopes, and the enhanced mafic composition compared to adjacent deposits.
 The pyroclastic deposits we identified in Montes Carpatus are similar in nature to previously identified Montes Carpatus pyroclastic deposits (Figure 17, locations C1–C4) [Gaddis et al., 1985] in that they appear as somewhat irregular, patchy dark-toned areas among superposed brighter-toned Copernicus ray materials. This morphology is not consistent with a model where each dark deposit erupted from a single localized vent. These exposures are therefore likely part of more widespread, overlapping deposits that have been partially obscured by the highland debris. The two locations near Gay-Lussac within the massifs of Montes Carpatus (Figure 20) could be exposures of pyroclastic material emplaced around the perimeter of Mare Imbrium from vents within the basin, similar to previously identified deposits such as those at Mons Huygens [Hawke et al., 1979]. The locations farther west near Tobias Mayer (Figure 19) are associated with smooth effusive deposits as well as features such as irregular craters, depressions, and domes that are also likely volcanic in origin. Similar to the Birt E deposit discussed above, these pyroclastic deposits could be late-stage volcanic materials erupted after emplacement of the regional mare basalt flows.
4. Conclusions and Future Work
 Our examination of new LROC WAC images of 47 low-albedo deposits resulted in identification of 12 new locations of probable pyroclastic deposits not previously cataloged. Newly identified pyroclastic deposits were generally found in settings similar to previously identified deposits, for example at the margins of major maria and within floor-fractured highland craters. However, a significant new finding is the discovery of potential localized pyroclastic deposits within floor-fractured craters Anderson E and F and adjacent to nearby Buys-Ballot on the lunar farside, isolated from other known deposits. The presence of volcanic deposits within this region of relatively thick crust is possibly the result of localized crustal thinning related to the Freundlich-Sharonov basin. This evidence of volcanic activity in a region generally devoid of volcanism provides additional constraints for models of the composition, structure, and thermal history of the crust and upper mantle in this region of the farside.
 Many of the 47 potential locations screened were eliminated from consideration based on inconclusive evidence regarding their mode of emplacement. Additional optical imaging, or analyses of other data sets such as radar, imaging spectroscopy, or thermal inertia, could result in identification of additional pyroclastic deposits, especially lighter-toned deposits. However, our search also confirms that most major regional and localized pyroclastic deposits have likely been identified on the Moon down to ∼100 m/pix resolution, and that additional newly identified dark-toned deposits are likely to be either isolated small deposits or additional portions of discontinuous, patchy deposits. Based on the locations where we identified previously unidentified pyroclastic deposits, the greatest potential for identification of additional pyroclastic deposits is likely to be in regions with other volcanic constructs associated with mare deposits (such as near Tobias Mayer west of Montes Carpatus), highland locations along the margins of maria (such as near Gay-Lussac north of Copernicus), and smaller floor-fractured craters that have not yet been thoroughly imaged at higher resolution, particularly on the farside (such as at Anderson E and F).
 A common theme among these new and previously identified pyroclastic deposits is their frequent association with effusive lava flows. It is possible that future, more detailed examination of many volcanic deposits in LRO data and future improved data sets will find that hybrid deposits exhibiting elements of both effusive and pyroclastic emplacement are common, and perhaps even the norm, on the Moon. Testing this hypothesis will yield important insights into lunar magmatic and volcanic processes, particularly the role of volatiles in various styles of lunar volcanism. We plan to examine this issue by conducting detailed studies of a number of pyroclastic deposits incorporating multispectral WAC data for compositional assessment, high-resolution NAC data for morphologic studies, and topographic data, in order to assess the frequency with which pyroclastic and effusive deposits occur together as well as the geologic settings under which this is most common. Additional future work will focus on characterization of the compositional diversity of localized pyroclastic deposits and groups of deposits and examination of the relationship between deposit characteristics and geologic setting. Analysis of combined data sets involving LROC NAC and WAC, LOLA [Smith et al., 2010], MiniRF [Nozette et al., 2010], and DIVINER [Paige et al., 2010], as well as relevant data from the Chandrayaan-1 Moon Mineralogy Mapper (M3) [Green et al., 2007; Staid et al., 2011] and the Kaguya mission [Kato et al., 2006] will provide an unprecedented opportunity to examine the detailed characteristics of these deposits. Such studies will help determine the relationship between the source regions for co-located effusive and pyroclastic deposits and will advance efforts to relate localized pyroclastic deposit characteristics (e.g., areal extent, volume, composition, and vent configuration) to the geologic setting within which they occur.
 This research was supported by a grant to J.F.B. from the NASA Lunar Reconnaissance Orbiter participating scientist program. We extend our deepest thanks to the capable men and women on the LROC operations and science teams, and on the LRO Project in general, for their outstanding work in collecting new NAC and WAC images of these and other important lunar geologic features.