Global distribution of lunar impact melt flows

Icarus 239 (2014) 105–117

Contents lists available at ScienceDirect

Icarus journal homepage: www.elsevier.com/locate/icarus

Global distribution of lunar impact melt flows C.D. Neish a,⇑, J. Madden b, L.M. Carter c, B.R. Hawke d, T. Giguere d,e, V.J. Bray f, G.R. Osinski g, J.T.S. Cahill h a

Department of Physics and Space Sciences, Florida Institute of Technology, Melbourne, FL 32901, United States Franklin and Marshall College, Lancaster, PA 17603, United States c NASA Goddard Space Flight Center, Greenbelt, MD 20771, United States d University of Hawai’i at Manoa, Honolulu, HI 96822, United States e Intergraph Corporation, Box 75330, Kapolei, HI 96707, United States f Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ 85721, United States g Centre for Planetary Science and Exploration, Departments of Earth Sciences and Physics and Astronomy, University of Western Ontario, London, Ontario N6A 3K7, Canada h The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, United States b

a r t i c l e

i n f o

Article history: Received 23 December 2013 Revised 29 May 2014 Accepted 31 May 2014 Available online 12 June 2014 Keywords: Impact processes Cratering Moon Moon, surface

a b s t r a c t In this study, we analyzed the distribution and properties of 146 craters with impact melt deposits exterior to their rims. Many of these craters were only recently discovered due to their unusual radar properties in the near-global Mini-RF data set. We find that most craters with exterior deposits of impact melt are small, 620 km, and that the smallest craters have the longest melt flows relative to their size. In addition, exterior deposits of impact melt are more common in the highlands than the mare. This may be the result of differing target properties in the highlands and mare, the difference in titanium content, or the greater variation of topography in the highlands. We find that 80% of complex craters and 60% of simple craters have melt directions that are coincident or nearly coincident with the lowest point in their rim, implying that pre-existing topography plays a dominant role in melt emplacement. This is likely due to movement during crater modification (complex craters) or breached crater rims (simple craters). We also find that impact melt flows have very high circular polarization ratios compared to other features on the Moon. This suggests that their surfaces are some of the roughest material on the Moon at the centimeter to decimeter scale, even though they appear smooth at the meter scale. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Flow-like deposits of smooth, low albedo material are observed on the Moon, typically around young, fresh craters (Shoemaker et al., 1968; El-Baz, 1972). These flows are interpreted to be impact melt, mixtures of clasts and melted material that are emplaced during the late stages of impact crater formation (Howard and Wilshire, 1975; Hawke and Head, 1977; Osinski et al., 2011). Melt generation is a fundamental part of the impact process and is influenced by a number of factors including impact velocity, impact angle and the composition of the impactor and the target (e.g., Grieve et al., 1977; Osinski et al., 2012). Thus, the origin and emplacement of impact melt flows may provide important insights into the impact cratering process in general. Hawke and Head (1977) noted that melt distribution patterns on crater exteriors tend to be asymmetric and may relate to the pre-impact topography. This led Hawke and Head (1977) to

⇑ Corresponding author. E-mail address: cneish@fit.edu (C.D. Neish). http://dx.doi.org/10.1016/j.icarus.2014.05.049 0019-1035/Ó 2014 Elsevier Inc. All rights reserved.

suggest that impact melt generated in the transient cavity of complex craters is moved upward and outward during the modification stage, as wall slumping and rebound proceed, resulting in melt movement over topographic lows in the rim crest. Osinski (2004) found evidence for similar movement of impact melt out of Ries crater on Earth (D = 24 km). In this analysis, Osinski (2004) suggested that the Ries flow occurred during the modification stage of crater formation, as movements associated with the central uplift imparted an outward-directed flow to the melt. For simple craters, previous data sets show little evidence for exterior flow of impact melt around craters with D < 10 km (Howard, 1974; Hawke and Head, 1977). Cintala and Grieve (1998) attributed this observation to a relatively small amount of melt becoming choked with an abundance of cold clasts, increasing the melt’s viscosity and chilling it rapidly. However, new data from the Lunar Reconnaissance Orbiter (LRO) suggests that flows can form around craters as small as 600 m in diameter (Stopar et al., 2014, submitted for publication). Osinski et al. (2011) suggest that such flows should be possible, but primarily for oblique impacts, where the sub-surface flow field is displaced downrange (Pierazzo

106

C.D. Neish et al. / Icarus 239 (2014) 105–117

and Melosh, 2000; Anderson et al., 2004), or impacts into varied topography, which frequently form craters with breached rims. In this work, we seek to quantify the role that pre-existing topography and impact direction play in the ejection of impact melt from simple and complex craters, updating the analysis completed by Hawke and Head (1977) with new data. The launch of LRO in June 2009 has afforded new views of lunar impact melt flows, allowing for a reassessment of their global distribution. Lunar impact melts have been primarily studied at optical wavelengths (Hawke and Head, 1977; Bray et al., 2010; Osinski et al., 2011; Plescia and Cintala, 2012; Denevi et al., 2012; Stopar et al., 2014, submitted for publication), but complementary information can be obtained by observing impact melts at radar wavelengths (Campbell et al., 2010; Carter et al., 2012). Indeed, since radar data is sensitive to surface and sub-surface roughness, it can highlight rough impact melts, even when they are not easily seen in optical data. This is because many of these melts are degraded, and partially buried by regolith, making them difficult to recognize in optical images (e.g., Fig. 1). One of the main advantages of radar is its ability to detect melt flows around Eratosthenian craters (1.1– 3.2 Ga), such as Gerasimovich D (Neish et al., 2011; Carter et al., 2012; Fig. 1) and Aristillus (Campbell et al., 2010). Orbital radars, such as Mini-RF on LRO, also maintain a consistent look angle, mitigating any issues related to lighting geometry (i.e., melts hidden in shadow at high incidence angles, or lacking textural information at low incidence angles). Radar data provides the capability to identify melt flows quickly and easily over large areas of the lunar surface. Some impact melts were identified in radar data on the lunar near side (Campbell et al., 2010), but most melts have yet to be

Fig. 1. The crater Gerasimovich D (22.3°S, 238.4°E) as seen at (a) visual wavelengths by the LROC WAC and (b) at S-Band (12.6 cm) by Mini-RF. The radar image is displayed as colorized circular polarization ratio over total radar backscatter. An arrow indicates the edge of a newly discovered impact melt flow (Neish et al., 2011; Kramer et al., 2011; Carter et al., 2012). In this figure, north is up. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

studied at radar wavelengths on the lunar far side, due to the lack of global radar data prior to the launch of the Mini-RF instrument on LRO (Nozette et al., 2010). Yet impact melts are globally important phenomena. Of the 56 impact melt flow bearing craters identified by Hawke and Head (1977), more than 50% were located on the lunar far side (31 out of 56). There are likely many exterior impact melt deposits on the far side of the Moon that have yet to be recognized in optical data, due to non-optimal lighting conditions or burial under a thin layer of regolith. In this work, we identified new impact melt flows in the MiniRF data set. Throughout the paper we will often refer to these external melt deposits as ‘flows’ as a general term for the veneers, ponds, and flow features seen exterior to craters, a precedent first set by Howard and Wilshire (1975). For each flow-like deposit inferred from radar images, we confirmed the presence of melt at the crater in overlapping Lunar Reconnaissance Orbiter Camera (LROC) Narrow Angle Camera (NAC) images (Robinson et al., 2010). We then determined the topography of the crater using Lunar Orbiter Laser Altimeter (LOLA) (Smith et al., 2010) or LROC Wide Angle Camera (WAC) stereo data (Scholten et al., 2012), and where applicable, the impact direction from Clementine UVVIS data (Eliason et al., 1999). This new survey provides a more complete global picture of lunar impact melt flows, allowing us to address fundamental questions about the emplacement of melt during the impact cratering process.

2. Observations We examined the Mini-RF images for morphologies consistent with melt flow features using global, 100 m/pixel mosaics recently produced by Cahill et al. (2014, submitted for publication). These mosaics include all of the S-Band (12.6 cm) monostatic data acquired during the LRO mission, and cover close to two-thirds of the lunar surface. In actuality, our effective sampling area is somewhat larger than that, since only a portion of the crater needs to be seen in one of the long, thin radar swaths. Once the crater is identified as having possible exterior impact melt deposits, a more complete view of the deposits can be assessed by examining overlapping optical data. Due to the gaps in the Mini-RF coverage, a few large craters with exterior melt deposits may have been missed in this survey. However, given the advantages of using radar data to identify melt flows, we contend that this represents the best global survey possible with the present data set. Global mosaics were produced for all four Stokes parameters (S1, S2, S3, and S4), as well as for daughter products such as the circular polarization ratio (CPR). CPR is one of the most useful indicators of surface roughness, making this an effective tool for identifying impact melts. This value is defined as the ratio of the backscattered power in the same-sense circular polarization as was transmitted (SC) to the opposite-sense circular polarization (OC). When an incident circularly polarized radar wave is backscattered off an interface, the polarization state of the wave changes. Thus flat, mirror-like surfaces, dominated by single-bounce reflections, tend to have high OC returns and low CPR values. Rough surfaces, dominated by multiple-bounce reflections, tend to have roughly equal OC and SC returns, with CPR values approaching unity. In some cases, the presence of abundant dipole or dihedral (corner) reflectors can lead to CPR values that exceed unity (i.e., CPR > 1) (Campbell, 2012). Impact melt deposits typically appear bright in radar images and have CPR values up to and exceeding unity, making them relatively easy to distinguish from the background terrain (Campbell et al., 2010; Carter et al., 2012; Fig. 1). The radar brightness indicates increased surface and/or near subsurface roughness close to the scale of the radar wavelength (in the case of Mini-RF, centimeter

107

C.D. Neish et al. / Icarus 239 (2014) 105–117

Table 1 Seventy-five lunar impact melt deposits exterior to their crater rim identified in the Mini-RF data set. Distances in italics indicate the measurement was made with radar data; distances in bold indicate the measurement was made with optical data. Crater

D (km)

Latitude (°)

Longitude (°E)

Max distance of melt from rim (crater radii)

Most extensive melt deposits

Most extensive ejecta

Topographically low rim crest

SE of Olcott S Rim of Birchoff K W of Isaev SE of Coriolis G W of Riccius E of Polybius R N of Nusl E Herigonius K In Weyl #2 SW of Maksutov SE Rim of Leavitt N of Tsinger N of Healy J NW of Compton NE of Davisson E of Lodygin S of Nassau Schroter D SW of Tiling NE of Compton Rim of Ley S of Kolhorster S of Tsinger NE of Orlov D NW of Galois B S of Hertzsprung Inside Pingre Eimmart A NE of Langevin C Rim of O’Day Stevinus A S Rim of Donner S of Dirichlet In Alder E of Lenz W rim of Fizeau Messier S of Schonfeld NE of Schrodinger E of Malyy Lagrange D Rim of Virtanen Furnerius A N of Chebyshev W of Hutton Pa SW of Artem’ev L S of Alter Messier A Pythagoras K SE of Pavlov N of Zhukovsky Gauss J E of Michelson Bessel Whipple Janssen K SE of Chalonge Ventris M NW of Kohlschotter von Bekesy F Slipher S Moore F Abbe H Stefan L Gerasimovich D Thales Petavius B Plante Das Newcomb Helmholtz D Ohm

1.5 2.2 2.4 2.5 2.8 2.9 3.0 3.0 3.1 3.3 3.6 3.8 4.0 4.0 4.4 4.7 4.8 5.0 5.3 5.8 6.0 6.0 6.3 6.5 6.6 6.8 7.0 7.0 7.5 7.5 7.5 7.5 7.6 8.0 9.0 9.0 10.0 10.0 10.0 11.0 11.0 11.0 11.0 11.0 11.6 11.7 12.0 12.0 12.0 13.0 13.5 14.0 15.5 15.6 15.7 16.0 16.0 16.2 18.5 20.0 24.0 24.0 25.0 26.0 26.0 31.0 33.0 37.0 38.0 40.0 46.0 64.0

17.4 56.8 17.7 0.4 37.5 25.7 34.4 12.8 17.0 41.4 45.2 57.3 30.9 59.5 34.6 17.1 26.4 4.5 53.0 58 41.3 6.1 55.3 24.1 9.6 4.1 58.9 24.1 47.0 31.0 31.9 32.2 9.1 48.4 3.4 58.7 1.9 43.6 71.3 21.4 34.9 15.8 33.6 28.8 35.5 6.2 15.3 2.0 67.3 30.9 11.6 40.6 7.5 21.7 89.1 46.2 22.0 5.7 15.4 52.8 48.9 37.4 58.2 44.6 22.3 61.6 19.9 10.2 26.6 29.8 66.1 18.4

120.0 214.2 144.4 175.0 23.1 27.9 169.8 323.5 237.8 188.1 221.7 175.7 250.8 100.5 187.4 215.6 177.3 350.5 226.0 114.6 156.5 243.9 174.5 187.3 206.6 227.7 287.3 65.7 166.9 158.0 51.6 97.9 207.7 183.4 259.7 223.0 47.6 262.5 162.4 108.6 287.5 177.3 59.0 227.2 166.2 214.8 250.8 47.0 284.2 145.5 192.1 72.7 243.7 17.9 118.2 42.3 244.5 157.9 151.6 137 158.7 185.0 177.9 252.3 238.4 50.2 57.1 163.3 223.2 43.7 54.1 246.5

1.3 2.9 0.7 0.9 3.4 3.3 1.3 1.7 0.6 1.0 0.7 1.0 1.5 2.4 2.0 0.3 4.0 1.4 0.6 5.2 1.3 0.3 0.9 1.3 1.4 0.2 0.4 1.5 0.3 0.9 1.4 3.8 0.3 2.1 2.6 2.4 2.0 1.6 2.6 0.2 1.4 1.1 0.3 1.6 0.9 1.0 1.5 5.3 0.3 1.0 0.2 1.5 1.0 1.3 2.4 1.6 1.2 1.8 0.7 0.3 2.0 1.3 0.7 1.1 2.0 1.0 1.0 1.4 1.0 1.0 0.5 0.4

SW SW S, SE, SW SW SW S NW NE NW S E NW E SW S E SE NW NE SE SE E W NW, SW NW NE S N N NE NE N SW SW NW E W W WSW SSE NE W NW S, NE W, WNW SE SW W NE W, NW NW SE NE ENE, SE S W N N W SW, NE NE SE W N W NE SE NW E, NW, S? N SE SE

E – – – SW NNW – NE SE? – – – – – ? NE? SW? – E? E – – – – E W? – – N – ESE? – – ? – – W – NE – – – NW? – – – – W – NW ? – – – – – – – – – – – NW – – – SSW – – – – SW

SE SE N SSW N NE SSW S NW NE W NW E SE SE E SW NW, SW ENE SE NNW SE SW NW NW S, NE W SE N NW NE N E W NW E W, E SW, NW NE N NE W WNW S, NE SW NE, S ESE E SE SE W SE NE ENE W W N N S SW NE SW SW N W SE ESE W E S E SE (continued on next page)

108

C.D. Neish et al. / Icarus 239 (2014) 105–117

Table 1 (continued)

a

Crater

D (km)

Jackson Hayn Vavilov

71.0 87.0 98.0

Latitude (°) 22.4 64.6 0.9

Longitude (°E)

Max distance of melt from rim (crater radii)

Most extensive melt deposits

Most extensive ejecta

Topographically low rim crest

196.9 83.9 138.8

0.4 1.2 0.7

W SSE ENE

SE – –

WNW SE ENE

This is a rare example of a radar dark impact melt flow, previously discussed in Carter et al. (2012).

to decimeter-scale roughness). For each possible impact melt deposit identified, we looked for evidence for melt in corresponding LROC NAC images. We used the criteria established in Hawke and Head (1977) to search for ‘‘. . .indications of fluid flow (flow lineations, leveed channels, ponding of material to a level surface).’’ Additional features associated with impact melt deposits, such as low albedo and cooling cracks in ponds and tension cracks in veneers can also be identified in these images (Howard, 1974; Howard and Wilshire, 1975; Hawke and Head, 1977). Given the resolution of the Mini-RF global mosaics (100 m/pixel), we limit ourselves to craters with diameters > 2 km. A comprehensive study of smaller impact melt bearing craters was conducted by Stopar et al. (2014, submitted for publication). We made no attempt to look specifically at craters of a specific age, to avoid any biases in our survey. However, we note that more than a third of the craters with D > 10 km in our sample were identified in Braden and Robinson (2013) as Copernican, supporting the idea that impact melt is often present around the youngest craters on the Moon. The survey conducted by Braden and Robinson (2013) was limited to those craters within ±40° of the equator, where crater rays can be identified, so more Copernican craters may be present in our sample at higher latitudes. If the crater was ‘confirmed’ to have exterior melt deposits, we documented the parent crater diameter and location, the flow

direction, and the maximum distance of the melt from the crater rim (Table 1, which includes our initial observations reported in Carter et al. (2012)). We also include several impact melts first identified and reported in LROC publications (Table 2) and the list of impact melt deposits identified prior to the LRO mission, from Howard and Wilshire (1975), Hawke and Head (1977) and Campbell et al. (2010) (Table 3). For these craters, impact melt direction and distance have been updated using modern spacecraft data. Where possible, the maximum distance of the melt from the crater rim was measured in the Mini-RF images. If radar data was not available at the terminus of the melt deposits, we used LROC NAC or WAC data (depending on the size of the crater and the availability of the data). All distance measurements were conducted on map-projected images using tools made freely available with the USGS ISIS program. Since radar is capable of sensing buried flows and identifying flows under non-optimal lighting conditions, the radar length is often longer than the corresponding optical length (e.g., Figs. 2 and 3). Thus, our average results represent a lower limit on the length of impact melt flows (since the data is a combination of the ‘shorter’ optical lengths and the ‘longer’ radar lengths). Several examples of newly identified impact melt deposits are shown in Figs. 2–8. For the compiled data set, we used LOLA and/or WAC stereo topography data to determine the direction of rim crest low for

Table 2 Fourteen lunar impact melt deposits exterior to their crater rim identified in LROC publications. Distances in italics indicate the measurement was made with radar data; distances in bold indicate the measurement was made with optical data. Longitude (°E)

Max distance of melt from rim (crater radii)

Most extensive melt deposits

Most extensive ejecta

Topographically low rim crest

Ref.

24.2

209.9

0.75

N

–

NE

2.7

18.8

273.5

–

NW

–

S, NW

E of Atlas

2.9

46.7

49.8

3.70

NE, NW

downrange

NE, S

SW of Steno N

3.0

30.0

161.0

2.30

NNW

downrange

NE

Hesiodus E

3.0

27.9

344.6

0.30

NE

–

N, S

Inside Yablochkov

3.5

60.9

126.8

2

S

–

E

Lichtenberg B

4.8

33.3

298.5

0.40

SW

–

NW

SE of Stein C

5.0

7.9

183.0

0.40

SW

NW?

NE

Rim of Lowell

9.0

13.3

257.6

3

NW

–

NW

W Rim of Joule T

13.8

27.7

211.4

1.3

SE

SE?

SE

SE Rim of Paraskevopoulos S N Rim of Korolev X

15.0

47.9

205.7

1.3

NW

W?

NW

16.4

1.1

200.5

1.50

S

–

S

Byrgius A

18.0

24.5

296.3

1.70

NE, (W)

–

W

Giordano Bruno

22.0

35.9

102.8

1.00

WSW

–

W

Denevi et al. (2012) Osinski et al. (2011) Hawke et al. (2011) Hawke et al. (2011) Denevi et al. (2012) Koeber et al. (2012) Denevi et al. (2012) Osinski et al. (2011) Koeber et al. (2012) Koeber et al. (2012) Koeber et al. (2012) Osinski et al. (2011) Hawke et al. (2010) Bray et al. (2010)

Crater

D (km)

Latitude (°)

S of Ingalls G

2.4

NE Rim of Einstein

Note: This list does not include the numerous impact melts found around small craters (D < 5 km) compiled by Stopar et al. (2014, submitted for publication).

109

C.D. Neish et al. / Icarus 239 (2014) 105–117

Table 3 Fifty-seven lunar impact melt deposits exterior to their crater rim identified prior to the LRO mission, from Howard and Wilshire (1975), Hawke and Head (1977), and Campbell et al. (2010). Major updates to the list compiled by Hawke and Head (1977) using modern spacecraft data are noted in bold. In almost all cases, the melt is observed further from the crater rim than previously noted.

a b c

Cratera

D (km)

Latitude (°)

Longitude (°E)

Max distance of melt from rim (crater radii)

Most extensive melt deposits

Most extensive ejecta

Topographically low rim crest

W wall of Lobachevskiy rim of Gibbs S Rim of Koval’sky Y near McKellar Römer Y NE of Meshchersky near Curie NW rim of Sklodowska W rim of Papaleski Chauvenet L O’Day M Aitken A SW rim of Cyrano A Sklodowska J Mandel’shtam F Dawes Epigenes A W of Fermi Lalande Koval’skiy P Becvar Q Proclus Al-Khwarizmi K Saha E Inside Wiener F Zhukovskiy Z Green M Necho Stearns Mairan Aristarchus Olbers A / Glushko Crookesb Rutherfurd Anaxagoras Maunder Mandel’shtam R Aristillusc Cavalerius Zucchius O’Day Philolaus Von Neuman Sharonov King Fabricius Tycho Aristoteles Copernicus Theophilus Langrenus Pythagoras Hausen Petavius Tsiolkovskiy Humboldt Schrödinger

2.3 4.0 4.2 5.4 6.6 8.5 9.0 9.5 10.0 10.0 13.0 13.0 14.0 15.0 15.0 18.0 18.0 18.0 24.0 25.0 26.0 26.0 26.0 28.8 30.0 32.4 35.3 36.9 38.9 39.5 40.0 40.1 48.3 50.0 51.9 53.8 54.0 55.0 59.3 63.2 70.4 71.4 74.8 75.1 76.2 78.9 85.3 87.6 96.1 98.6 132.0 144.6 163.2 184.1 184.4 199.5 316.4

9.7 17.5 21.1 16.7 25.8 13.2 23.7 16.9 10.85 13.3 31.6 13.9 18.3 19.4 5.1 17.2 67 19.8 4.5 22.4 3.1 16.1 4.5 0.3 40.9 9.8 0.4 5.2 34.7 41.6 23.7 8.1 10.4 61.1 73.5 14.5 4.4 33.9 5.1 61.4 30.4 72.2 40.3 12.4 5 42.8 43.3 50.2 9.6 11.5 8.9 63.7 65.1 25.4 20.4 27 74.7

111.7 85.2 100.02 170 36.3 127.6 88 93.9 162.08 137.8 157.1 173.5 158.1 98 166.1 26.3 0.4 117.3 8.6 100.7 124.3 46.9 108.2 108 149.9 167.2 133.1 123.2 162.6 43.5 47.3 77.7 165.1 12.7 10.2 93.9 159.8 1.2 66.9 50.6 157.3 32.9 153.2 173.1 120.5 41.8 11.2 17.3 20.1 26.3 61 63 88.5 60.8 129 81 132.9

2.2 2.0 0.8 1.4 1.3 1.7 3.8 1.2 1.1 1.1 0.9 – – 0.5 0.6 0.8 1.0 0.9 0.5 0.2 0.3 1.0 1.0 0.5 1.0 0.8 0.1 1.0 1.4 0.4 0.8 1.8 0.7 1.8 0.9 0.2 0.9 2.0 0.1 0.3 1.5 1.2 0.2 0.2 0.8 0.3 0.7 0.5 0.6 0.9 0.7 0.9 1.2 0.6 1.0 1.0 0.7

E SSW SW E SSW SE NNW SSW N SSE N N NNE NW NW SE W NNE SE NNE ENE N SSW NNW N N NE NE SE SW E, S NNW SE W ESE W NNE NE NE E ESE E NE ENE NE S, WSW E S S NE SSE E WSW SW SE, S SE E

E SSW ESE SE SSW NW – SW – W – – NNE – NW SW SSW NNE N NE NE NE – – S, N – – N – – SE NE W? NNE ESE SSE – – ESE – SE NE – – NNW SW E N NW NE SE S – SSE SSE SSE NE

E SW NNW E SSW SE, N S SSW NW SE N NE NNE NW NE, NW NW W NNE ESE NNE ENE SW SSW NNW N NNW E NE SE SW E NNW SE W E SW NNE ENE N,NE E ESE SW NE S, SE NE S SW SE NW, S N, NE, E S SW WSW N W ENE S,E

The melt ‘NW of Valier’ has been removed from the original list of Hawke and Head (1977), as there is no evidence for melt in modern imagery. The melt deposits at Crookes were noted by Howard and Wilshire (1975), but not included in Hawke and Head (1977). The melt deposits at Aristillus were discovered by Campbell et al. (2010) using Earth-based radar.

each crater. Given the kilometer-sized gaps between the LOLA profiles near the lunar equator, we only used this data set for craters larger than 10 km in diameter (Smith et al., 2010). For smaller craters, we used the 100 m/pixel WAC stereo data set (Scholten et al., 2012). Note that the smallest diameter objects for which reliable height information can be obtained with this data set is 1.5 km, as a result of the matching masks used in the area-based stereo

approach (Scholten et al., 2012). This is within the range of crater diameters studied in this work (>2 km). If the crater’s ballistic ejecta blanket was visible, high-Sun (30° incidence, 0° emission), Clementine 750 nm images were used to determine impact direction. Oblique impacts with incidence angles less than 45° form a wedge-shaped ‘forbidden zone’ in the visible ejecta blanket uprange of the crater (Melosh, 1989). These data

110

C.D. Neish et al. / Icarus 239 (2014) 105–117

of melt flow at a 13 km crater southeast of Pavlov correlates exactly with the downrange direction of impact (Fig. 7). The large amount of melt at Messier and Messier A (Fig. 6) is a particularly unusual case, given that the craters were likely formed by an extremely oblique impact (<5°) (Gault and Wedekind, 1978). The volume of impact melt predicted for such an oblique impact is less than 10% of the volume expected in a vertical impact (Pierazzo and Melosh, 2000). The observation of melt at Messier may suggest that shear heating contributes to melting in the most oblique cases (Schultz, 1996).

3. Results and discussion 3.1. Global distribution of melt flow bearing craters

Fig. 2. The 87 km diameter crater Hayn (64.6°N, 83.9°E) has a well-defined impact melt flow seen in (a) LROC WAC and (b) Mini-RF images (indicated by a white arrow). Impact melts at larger, older craters are often degraded and covered by regolith, making them difficult to identify in optical data. The radar image is displayed as colorized circular polarization ratio over total radar backscatter. North is up in these images. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

allow us to determine the relative importance of pre-existing topography (e.g., Figs. 4 and 5) and impact direction (e.g., Figs. 6– 8) in impact melt emplacement. For example, the direction of the melt flow at a 9 km crater on the rim of Donner correlates exactly with a topographic low in the rim crest (Fig. 4), while the direction

The data compiled in this work represents a total of 146 craters with exterior deposits of impact melt. This increases the known population of impact melt flow bearing craters by more than 60% and allows us to reexamine the distribution and characteristics of impact melt deposits on the Moon. In general, the exterior impact melt deposits identified in this work are (a) more common in the highlands than the mare (Fig. 9) and (b) found around smaller craters (Fig. 10) than the craters identified by Hawke and Head (1977). Nelson et al. (2013) mapped the mare regions on the Moon using a combination of LROC WAC and Clementine UVVIS mosaics, and found that mare regions constitute 18.7% of the area between 60°S and 60°N. If we assume that the polar regions are almost entirely highlands regions, this equates to 16.2% of the whole Moon, similar to an earlier estimate of 17% by Head (1975). Of

Fig. 3. Whipple crater, located near the north pole (89.1°N, 118.2°E), seen here in the (a) LROC WAC and (b) Mini-RF polar mosaics. This crater has what appears to be an impact melt flow, indicated by a white arrow. The lighting conditions near the pole make confirmation of this melt flow difficult in optical imagery, but features consistent with cooling cracks are visible on the sunlit rim in the (c and d) LROC NAC 2 m/pixel polar mosaic (NAC_POLE_P892N1350.IMG) (black arrows). The location of (c) is shown as a white box in (a and b), and the location of (d) is shown as a white box in (c).

C.D. Neish et al. / Icarus 239 (2014) 105–117

Fig. 4. This 7.5 km diameter crater (32.2°S, 97.9°E) on the rim of the larger 58 km diameter Donner crater has an impact melt flow that extends four crater radii from the rim. Several indications of fluid flow, including lobate edges, leveed channels, and ponding of material, are obvious. The edge of the melt flow is difficult to discern in the (a) LROC NAC images (M1111165703LE.IMG, M115672963LE.IMG, M115672963RE.IMG, M131005361LE.IMG, M131005361RE.IMG), but is well defined in the (b) Mini-RF image. The flow direction corresponds exactly to the lowest point in the crater rim, as seen in (c) LOLA topography (displayed over total radar backscatter). The radar image is displayed as colorized circular polarization ratio over total radar backscatter. North is up in these images. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5. This 5.8 km diameter crater (top left; 58.0°N, 114.6°E) has one of the longest impact melt flows (relative to its size) identified in our survey, extending 5.2 crater radii away from its rim. White arrows identify portions of the melt flow in (a) a LROC NAC image mosaic (M185148771LE.IMG, M185148771RE.IMG, M187514823LE.IMG), (b) a Mini-RF image (colorized CPR over total radar backscatter, placed over LROC NAC images for context), and (c) LOLA topography. The melt direction corresponds to a topographically low portion of the crater rim, and the melt appears to have sufficient velocity to flow into and out of a nearby, smaller crater. North is up in these images. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

the 146 impact melt flow bearing craters studied in this work, at most 10 of them lie in mare regions (Lalande is a questionable case,

111

as it lies on a highlands/mare boundary). If such craters were evenly spread about the Moon, we would expect 24 of them to be in the mare (0.162  146). Thus, there is more than a factor of two discrepancy. Most exterior impact melt deposits are associated with Copernican or Eratosthenian craters, which post-date mare formation, so we do not expect the greater average age of the highlands to be the primary cause of the discrepancy. Several other explanations could account for the higher number of impact melt flow bearing craters in the highlands. The differing target properties of the highlands may enhance the production of impact melt during the cratering process. The regolith dominated highlands are more porous than the more coherent mare surfaces, and have thicker layers of megaregolith (Thompson et al., 2009; Kiefer et al., 2012; Wieczorek et al., 2013). Wünnemann et al. (2008) predict a modest increase in melt production with increasing porosity, because of the lower critical shock pressures required for melting in porous materials. Impact melt at terrestrial craters formed in porous rock (mixed sedimentary-crystalline and purely sedimentary targets) has historically been difficult to recognize (Osinski et al., 2008). However, revised melt estimates from several terrestrial craters (Osinski, 2004; Osinski et al., 2005, 2008) are consistent with model results of melt production in porous targets. Alternatively, the increased TiO2 content of the mare might obscure the detection of buried impact melt flows. Regolith with higher TiO2 content has a higher loss tangent, and as a result, the penetration depth of S-Band radar may be reduced to only 1–2 m (Campbell et al., 1997). Longer wavelength radars can penetrate deeper into the regolith, so future P-Band (70 cm) mapping may expand the population of impact melt flow bearing craters in the mare. Finally, there may simply be more exterior impact melt deposits in the highlands because there is greater topographic variability there. The varied topography of the highlands may provide more opportunities for melt to flow beyond the crater rim due to the greater chance of forming a breached rim (see, for example, Fig. 4). The melt would require less kinetic energy to overcome a smaller potential barrier. We note that the majority of craters in this study are smaller than 20 km in diameter (90 out of 146, or 62%). This is an unexpected result, since the limited amount of melt formed in small craters (D < 10 km) would likely be choked with cold clasts, increasing the melt’s viscosity and chilling it rapidly (Cintala and Grieve, 1998). Previous work suggested that small craters would have exterior melt deposits dominated by veneers that do not extend far from the crater rim (Hawke and Head, 1977). This new study shows the opposite – craters smaller than 10 km have the longest relative flow length of any size of crater on the Moon, with an average distance of 1.6 crater radii from the rim crest (Fig. 11). It could be that these melt flows are simply easier to identify, since smaller craters are more abundant and statistically more likely to be younger and less degraded. The relatively short run-out distance of melt outside of complex crater rims may also be due to the fact that terrace formation moves a lot of the initially ejected melt back inside the crater (relative to the final crater rim), whereas with simple craters any melt that is emplaced beyond the rim remains outside. This will reduce the complex crater melt distances in two ways: (1) there is now not as much melt remaining outside of the crater rim to form flows, and (2) the rim diameter has been increased by crater modification, decreasing the ratio of melt distance to crater radius. If porosity is a contributing factor to melt production, smaller craters may benefit more since a higher percentage of the target affected by the impact is the porous upper layers of the regolith. More likely, though, a combination of preexisting topography and oblique impact geometries aids in melt emplacement exterior to simple craters (Hawke and Head, 1977; Osinski et al., 2011). More detailed measurements of the volume of melt in such craters is needed to determine if these results are

112

C.D. Neish et al. / Icarus 239 (2014) 105–117

Fig. 6. (a) Messier A (2.0°S, 47.0°E) has an intriguing ejecta pattern that appears to be dominated by impact melt. Like other impact melts, it is radar bright in the (b) Arecibo/ Green Bank Telescope (e.g., Campbell et al., 2010) and (c) Mini-RF S-Band images. A white box in the (a) LROC WAC image shows the location of (d) a LROC NAC image (M190287137LE.IMG). Here, arrows indicate impact melt flow fronts. Given that Messier and Messier A were likely formed by an extremely oblique impact (<5°), the melt direction in this case is almost certainly associated with the impact direction. The radar images are displayed as colorized circular polarization ratio over same sense radar backscatter. North is up in these images. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 7. This 13 km diameter crater SE of Pavlov (30.9°S, 145.5°E) seen in (a) low-Sun (morphology enhanced) LROC WAC image, (b) high-Sun (albedo enhanced) Clementine 750 nm image, and (c) Mini-RF images has an impact melt deposit on the NW rim of the crater (see LROC NAC inset in (d) (M115360776LE.IMG)). The ejecta pattern seen in the Clementine image suggests that the direction of this melt flow is associated with the direction of the impact, i.e., from the SE to the NW. The radar images are displayed as colorized circular polarization ratio over same sense radar backscatter. North is up in these images. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

C.D. Neish et al. / Icarus 239 (2014) 105–117

113

Fig. 8. This 10 km diameter crater NE of the Schrödinger basin (71.3°S, 162.4°E) seen in (a) LROC WAC and (b) Mini-RF images has a large impact melt deposit on the WSW rim of the crater (see LROC NAC inset in (d) (M112884286LE.IMG)). This is unusual, as the direction of the melt is not correlated with the rim crest low, as seen in the (c) LOLA topography (white arrow), or the direction of impact, which given the ejecta pattern appears to have been to the ENE. The radar images are displayed as colorized circular polarization ratio over total radar backscatter. North is up in these images. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 9. The position of lunar impact craters that possess exterior melt deposits, plotted over a Clementine UVVIS 750 nm global mosaic. The size of the markers is scaled to crater diameter. Red markers are impact melts identified in Mini-RF data (Table 1), yellow markers are impact melts identified in LROC data (Table 2), and blue markers are impact melts identified in pre-LRO observations (Table 3). Lines of latitude are separated by 30° and lines of longitude are separated by 45°. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

consistent with present models of impact melt production (i.e., Plescia and Cintala, 2012). 3.2. Impact melt emplacement This study allows us to reexamine the relative role of pre-existing topography and impact direction in melt emplacement. We

separated the data set into simple craters (for this study, only those with D 6 10 km, a total of 58 craters) and complex craters (for this study, only those with D > 20 km, a total of 55 craters). We then determined the relative directions of the exterior impact melt deposits compared to (a) the rim crest low and (b) the inferred direction of impact. Note that the data set does contain many examples of craters between 10 and 20 km in diameter, but we

114

C.D. Neish et al. / Icarus 239 (2014) 105–117

Fig. 10. (a) The size distribution of the craters with exterior impact melt deposits in the original survey by Hawke and Head (1977) (dark gray) and the present work (light gray). Most melt flow bearing craters have diameters less than 20 km. Note that in some cases the two histograms do not line up because the crater’s diameter has been updated in the present work (e.g., Schrödinger). (b) The size distribution of craters with diameters greater than or equal to 20 km (Head et al., 2010). Smaller craters are more abundant than larger craters, which may help to explain their prevalence in our sample.

Fig. 11. The average maximum distance that melt deposits occur from their parent crater rim, expressed in terms of crater radii, plotted for several diameter intervals (2–10 km, 10–20 km, 20–50 km, 50–100 km, and 100–350 km). The dark gray histogram represents the original survey completed by Hawke and Head (1977). The light gray histogram represents the present work, using the data in Tables 1–3. The original survey indicated a slight increase in melt distance with crater size, while the present survey suggests the opposite trend.

excluded these from all subsequent analysis since they tend to be transitional between simple and complex craters, with rim slumping but no obvious central peaks (Howard, 1974; Pike, 1977). For complex craters, we find a strong correlation between the direction of impact melt deposits and the rim crest low. Eighty percent of complex craters have melt directions that are within 45° of the rim crest low azimuth, while 53% of complex craters have melt directions that are within a few degrees of this azimuth (e.g., Gerasimovich D, Fig. 1) (Fig. 12a). (The directions listed here are given

Fig. 12. The correlation of the flow direction of the exterior impact melt deposits with the direction of the rim crest low for (a) complex and (c) simple craters. The correlation of the flow direction of the exterior impact melt deposits with the inferred downrange direction of the impact for (b) complex and (d) simple craters. The number of craters for which data is available (Tables 1–3) is given by n.

relative to the center of the crater, i.e., NE is 45° away from E and 180° away from SW.) In contrast, 55% of complex craters have melt directions that are within 45° of the inferred downrange direction of the impact, and only 13% of complex craters have melt directions that are within a few degrees of this azimuth (e.g., Tycho) (Fig. 12b). It should be noted that impact direction is often difficult to determine, so this data set is based only on those craters that we could determine impact direction and is somewhat smaller (n = 31 vs. n = 55). These results support the original conclusion of Hawke and Head (1977), who found that 93% of complex craters had melt directions within 45° of the rim crest low azimuth, and 50% had melt directions within a few degrees of this azimuth. One interpretation for this observation is that impact melt is emplaced during the modification stage of crater formation (Hawke and Head, 1977; Osinski et al., 2011). Uplift during cavity modification imparts an outward momentum to the melt-rich lining of the transient cavity, resulting in flow over the lowest point in the crater rim. After that point, the movement of melt – which is a fluid – will be largely controlled by the topography of the target region. For simple craters, there is no clear correlation between the melt direction and the rim crest low or the direction of impact. Only 60% percent of simple craters have melt directions that are within 45° of the rim crest low azimuth, and 41% of simple craters have melt directions that are within a few degrees of this azimuth (e.g., rim of Donner, Fig. 4) (Fig. 12c). Similarly, only 52% of simple craters have melt directions that are within 45° of the inferred downrange direction of the impact, and 32% of simple craters have melt directions that are within a few degrees of this azimuth (e.g., Messier, Fig. 6) (Fig. 12d). Both of these effects likely play a role in impact melt emplacement in simple craters, depending on the topography of the region (e.g., impact into crater rim or flat plain?) and the relative obliquity of the impact (Osinski et al., 2011).

C.D. Neish et al. / Icarus 239 (2014) 105–117

115

Fig. 13. Glushko crater (8.1°N, 282.3°E), as viewed by (a) LROC WAC and (b) Arecibo/Green Bank Telescope S-Band radar (colorized DLP over total radar backscatter; Campbell et al., 2010). A fresh impact melt flow is seen to the northwest of the crater. It is radar bright and has a lower DLP than the surrounding regolith, suggesting surface scattering from a rough surface. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.3. Physical properties of impact melt deposits Impact melts have some of the highest circular polarization ratios of any material on the Moon, with CPRs frequently near or exceeding one (Campbell et al., 2010; Carter et al., 2012). This suggests that the melts are ‘rougher’ at the centimeter to decimeter scale than most types of terrestrial lavas, which have average CPRs of 0.2 (pahoehoe), 0.5 (a0 a), and 0.8–1.0 (blocky lavas) at similar incidence angles to Mini-RF (Campbell, 2002). It is notable that lavas on Mars (Harmon et al., 2012) and some impact melts on Mercury (Neish et al., 2013) also have high CPR. The cause of this increased CPR is unknown, but could be the result of either surface or subsurface scattering. For example, scattering from rock edges and cracks (dipole scattering) can produce CPRs between one and two, while scattering from natural corner reflectors (dihedral scattering) can yield CPRs greater than two (Campbell, 2012). Morphologically, lunar impact melts appear most like pahoehoe flows on the Earth (Bray et al., 2010), which tend to have very low CPR values. However, some types of pahoehoe, namely rubbly pahoehoe (Keszthelyi et al., 2004), have not been studied at radar wavelengths. The increased CPR could also be due to subsurface scattering from clasts or voids within the melt. This option may be less likely than scattering from a rough surface, given the large power losses expected in such a process. However, we cannot rule it out entirely because there is evidence that impact melts may have lower density, and hence lower loss tangents, than other lunar rocks. Some samples of impact melt acquired during the Apollo program have lower densities than mare basalts (2350–2600 kg/m3 vs. 3000– 3300 kg/m3 for mare basalts, Kiefer et al., 2012). If this is true for impact melts in general, the radar may be able to penetrate more deeply into the melt than a typical basaltic flow. The penetration depth of a radar signal is dependent on its illuminating wavelength, k, the loss tangent of the substrate, tan d, and the real dielectric constant of the substrate, e0 (Campbell and Campbell, p 2006). To a first order, the probing length is given by k/(2p e0 tan d). Both the loss tangent and the real dielectric constant are dependent on the density of the substrate. For a density of 2500 kg/m3, the real dielectric constant is e0 = 1.96q = 5.4, where q is given in g/cm3 (Carrier et al., 1991; Campbell, 2002). The imaginary part of the dielectric constant, e00 , used to calculate the loss tangent tan d = e00 /e0 , is more difficult to measure, but appears to be more strongly dependent on the composition of the rock than its density. At a density of 2500 kg/m3 and a wavelength of 19 cm, e00 ranges from 0.01 to 0.2 for igneous rocks (Ulaby et al., 1988), leading to a penetration depth of 20–500 cm at S-Band (there is no corresponding data at 12.6 cm, but e00 tends to increase

with decreasing wavelength in this range (Carrier et al., 1991), which would tend to decrease the penetration depth). Thus, it is possible that S-Band radar could penetrate tens of centimeters into impact melt deposits, scattering off any voids or higher density clasts present there. One simple way to test the relative importance of surface vs. subsurface scattering is to determine the degree of linear polarizap tion (DLP) of the material, calculated by DLP = (S22 + S23)/S1. Targets with little subsurface scattering have low DLP, targets with diffuse subsurface scattering have moderate DLP, and targets with buried interfaces have high DLP (see Fig. 1 in Carter et al., 2011). To distinguish between these scenarios, we calculated the DLP for Glushko crater (Fig. 13). Glushko has a fresh impact melt flow (Campbell et al., 2010; Neish et al., 2013) that appears to be covered by very little to no regolith (unlike other melt flows, such as that at Gerasimovich D, shown in Fig. 1), so any scattering we see is from the impact melt flow surface, not a buried interface (which would tend to increase the DLP). We see that Glushko and its corresponding melt flow have a lower DLP than the surrounding regolith. Coupled with its radar bright appearance, we can infer that the impact melt flow is dominated by surface scattering off of a rough surface. Thus, it seems likely that impact melt flows are radar bright primarily due to their surface texture at a scale of centimeters to decimeters. This ‘microroughness’ is well below the resolution limit of the LROC NAC (10 cm vs. 50 cm), so even melts that look ‘smooth’ in NAC images may possess this surface texture. In the case of Glushko, its CPR at P-Band (70 cm) drops by only 10% compared to the CPR at S-Band, indicating that the surface is rough at multiple scales (Campbell et al., 2010). Lunar impact melts are clearly very complex surfaces, and additional work is needed to characterize lava flows (e.g., rubbly pahoehoe) and other rugged surfaces on the Earth that could produce CPRs similar to those of the lunar impact melt flows.

4. Conclusions In this study, we analyzed the distribution and properties of 146 craters with exterior deposits of impact melt. Many of these impact melt deposits were only discovered due to their unusual radar properties in the near-global Mini-RF data set. Contrary to previous studies, we find that most craters with exterior deposits of impact melt are small, 620 km in diameter, and that the smallest craters have the longest melt flows relative to their size. This suggests that a combination of pre-existing topography and oblique impact geometries greatly aids in the emplacement of melt exterior to

116

C.D. Neish et al. / Icarus 239 (2014) 105–117

simple craters. In addition, exterior deposits of impact melt are more common in the highlands than the mare (depleted by more than a factor of two in the mare). This may be the result of differing target properties in the highlands and mare, the difference in titanium content, or the greater variation of topography in the highlands. We also find that the vast majority of complex craters have exterior melt directions that lie within 45° of the lowest point in their rim. This suggests that melt is being driven up and over the crater rim during the modification stage of crater formation, when uplift imparts outward momentum to the melt lining the transient crater cavity (cf., Osinski et al., 2011). Melt emplacement in simple craters appears to involve several different mechanisms, and may occur in oblique impacts, where the sub-surface flow field is displaced downrange, or impacts into varied topography, which form breached rims. Impact melt deposits also appear to be some of the roughest material on the surface of the Moon. We suggest that this roughness is due to dipole or dihedral scattering from a surface that is rough on a scale of centimeters to decimeters, but smoother on a scale of meters (as viewed by the LROC NAC). Additional work is needed to find terrestrial analogues that display the same morphology and radar scattering properties of lunar impact melt flows. Acknowledgments We thank the LRO project for their effort in returning the data presented here. We also wish to thank Bruce Campbell and Lillian Ostrach for their careful reviews, and Natalie Glines for help in the early stages of this work. This work was supported in part by an appointment to the NASA Postdoctoral Program at the Goddard Space Flight Center, administered by Oak Ridge Associated Universities through a contract with NASA to C.N. References Anderson, J.L.B., Schultz, P.H., Heineck, J.T., 2004. Experimental ejection angles for oblique impacts: Implications for the subsurface flow-field. Meteorit. Planet. Sci. 39, 303–320. Braden, S.E., Robinson, M.S., 2013. Relative rates of optical maturation of regolith on Mercury and the Moon. J. Geophys. Res. 118, 1903–1914. Bray, V.J. et al., 2010. New insight into lunar impact melt mobility from the LRO camera. Geophys. Res. Lett. 37, L21202. Cahill, J.T.S., Thomson, B.J., Patterson, G.W., Bussey, D.B.J., Neish, C.D., Lopez, N.R., Turner, F.S., Aldridge, T., McAdam, M., Meyer, H.M., Raney, R.K., Carter, L.M., Spudis, P.D., Hiesinger, H., Pasckert, J.H., 2014. The miniature radio frequency instrument’s (Mini-RF) global observations of Earth’s Moon. Icarus, submitted for publication. Campbell, B.A., 2002. Radar Remote Sensing of Planetary Surfaces. Cambridge University Press, Cambridge, UK. Campbell, B.A., 2012. High circular polarization ratios in radar scattering from geologic targets. J. Geophys. Res. 117, E06008. Campbell, B.A., Campbell, D.B., 2006. Regolith properties in the south polar region of the Moon from 70-cm radar polarimetry. Icarus 180, 1–7. Campbell, B., Hawke, B., Thompson, T., 1997. Regolith composition and structure in the lunar maria: Results of long-wavelength radar studies. J. Geophys. Res. – Planets 102, 19307–19320. Campbell, B.A., Carter, L.M., Campbell, D.B., Nolan, M., Chandler, J., Ghent, R.R., Hawke, B.R., Anderson, R.F., Wells, K., 2010. Earth-based 12.6-cm wavelength radar mapping of the Moon: New views of impact melt distribution and mare physical properties. Icarus 208, 565–573. Carrier, W.D., Olhoeft, G.R., Mendell, W., 1991. Physical properties of the lunar surface. In: Heiken, G. (Ed.), Lunar Sourcebook. Cambridge University Press, Cambridge, UK, pp. 475–566. Carter, L.M., Campbell, D.B., Campbell, B.A., 2011. Geologic studies of planetary surfaces using radar polarimetric imaging. Proc. IEEE 99, 770–782. Carter, L.M. et al., 2012. Initial observations of lunar impact melts and ejecta flows with the Mini-RF radar. J. Geophys. Res. 117, E00H09. Cintala, M.J., Grieve, R.A.F., 1998. Scaling impact-melt and crater dimensions: Implications for the lunar cratering record. Meteorit. Planet. Sci. 33, 889–912. Denevi, B.W., Koeber, S.D., Robinson, M.S., Garry, W.B., Hawke, B.R., Tran, T.N., Lawrence, S.J., Keszthelyi, L.P., Barnouin, O.S., Ernst, C.M., Tornabene, L.L., 2012. Physical constraints on impact melt properties from Lunar Reconnaissance Orbiter Camera images. Icarus 219, 665–675.

El-Baz, F., 1972. King Crater and Its Environs. Apollo 16: Preliminary Science Report. NASA SP-315, C-3 pages, Published by NASA, Washington, DC, 1972, pp. 29–62. Eliason, E. et al., 1999. Mission to the Moon: The Clementine UVVIS Global Lunar Mosaic, PDS Volumes USA_NASA_PDS_CL_4001 through 4078, Produced by the U.S. Geological Survey and Distributed on CD Media by the Planetary Data System. Gault, D., Wedekind, J., 1978. Experimental studies of oblique impact. Proc. Lunar Sci. Conf. 9, 3843–3875. Grieve, R.A.F., Dence, M.R., Robertson, P.B., 1977. Cratering processes: As interpreted from the occurrences of impact melts. In: Roddy, D.J., Pepin, R.O., Merrill, R.B. (Eds.), Impact and Explosion Cratering. Pergamon Press, New York, NY, pp. 791– 814. Harmon, J.K., Nolan, M.C., Husmann, D.I., Campbell, B.A., 2012. Arecibo radar imagery of Mars: The major volcanic provinces. Icarus 220, 990–1030. Hawke, B.R., Head, J.W., 1977. Impact melt in lunar crater interiors. In: Roddy, D.J., Pepin, R.O., Merrill, R.B. (Eds.), Impact and Explosion Cratering. Pergamon Press, New York, NY, p. 815. Hawke, B.R. et al., 2010. Byrgius a crater impact melts—An LROC perspective. Lunar Planet. Sci. 41, 1611. Hawke, B.R. et al., 2011. A tale of two craters: Impact melt at two very small craters on the Moon. Lunar Planet. Sci. 42, 2347. Head, J.W., 1975. Lunar mare deposits: Areas, volumes, sequence, and implication for melting in source areas. In: Proceedings of the Conference on Origins of Mare Basalts and their Implications for Lunar Evolution, LPI Contribution 234, pp. 66–69. Head, J.W. et al., 2010. Global distribution of large lunar craters: Implications for resurfacing and impactor populations. Science 329, 1504–1507. Howard, K.A., 1974. Fresh lunar impact craters – Review of variations with size. Proc. Lunar Sci. Conf. 5, 61–69. Howard, K.A., Wilshire, H.G., 1975. Flows of impact melt at lunar craters. J. Res. U.S. Geol. Survey 3, 237–251. Keszthelyi, L. et al., 2004. Icelandic analogs to martian flood lavas. Geochem. Geophys. Geosyst. 5, Q11014. Kiefer, W.S., Macke, R.J., Britt, D.T., Irving, A.J., Consolmagno, G.J., 2012. The density and porosity of lunar rocks. Geophys. Res. Lett. 39, L07201. Koeber, S.D., Denevi, B.W., Stopar, J.D., Robinson, M.S., 2012. Lunar craters with exterior impact melt deposits. NASA Lunar Science Forum, vol. 5. Kramer, G.Y. et al., 2011. M3 spectral analysis of lunar swirls and the link between optical maturation and surface hydroxyl formation at magnetic anomalies. J. Geophys. Res. 116, E00G18. Melosh, H.J., 1989. Impact Cratering. Oxford University Press, Oxford, UK. Neish, C.D., Blewett, D.T., Bussey, D.B.J., Lawrence, S.J., Mechtley, M., Thomson, B.J., 2011. The surficial nature of lunar swirls as revealed by the Mini-RF instrument. Icarus 215, 186–196. Neish, C.D., Blewett, D.T., Harmon, J.K., Coman, E.I., Cahill, J.T.S., Ernst, C.M., 2013. A comparison of rayed craters on the Moon and Mercury. J. Geophys. Res. – Planets 118, 2247–2261. http://dx.doi.org/10.1002/jgre.20166. Nelson, D.M., Koeber, S., Robinson, M., 2013. Mapping lunar mare with LROC and Clementine image data. NASA Lunar Science Forum, vol. 6. Nozette, S. et al., 2010. The Lunar Reconnaissance Orbiter Miniature Radio Frequency (Mini-RF) technology demonstration. Space Sci. Rev. 150, 285–302. Osinski, G.R., 2004. Impact melt rocks from the Ries structure, Germany: An origin as impact melt flows? Earth Planet. Sci. Lett. 226, 529–543. Osinski, G.R., Lee, P., Parnell, J., Spray, J.G., Baron, M., 2005. A case study of impactinduced hydrothermal activity: The Haughton impact structure, Devon Island, Canadian High Arctic. Meteorit. Planet. Sci. 40, 1859–1877. Osinski, G.R., Grieve, R.A.F., Collins, G.S., Marion, C., Sylvester, P., 2008. The effect of target lithology on the products of impact melting. Meteorit. Planet. Sci. 43, 1939– 1954. Osinski, G.R., Tornabene, L.L., Grieve, R.A.F., 2011. Impact ejecta emplacement on terrestrial planets. Earth Planet. Sci. Lett. 310, 167–181. Osinski, G.R., Grieve, R.A.F., Marion, C.L., Chanou, A., 2012. Impact melting. In: Osinski, G.R., Pierazzo, E. (Eds.), Impact Cratering: Processes and Products. Wiley-Blackwell, Chichester, pp. 125–145. Pierazzo, E., Melosh, H.J., 2000. Melt production in oblique impacts. Icarus 145, 252– 261. Pike, R.J., 1977. Size-dependence in the shape of fresh impact craters on the Moon. In: Roddy, D.J., Pepin, R.O., Merrill, R.B. (Eds.), Impact and Explosion Cratering. Pergamon Press, New York, NY, p. 489. Plescia, J.B., Cintala, M.J., 2012. Impact melt in small lunar highland craters. J. Geophys. Res. 117, E00H12. Robinson, M.S. et al., 2010. Lunar Reconnaissance Orbiter Camera (LROC) instrument overview. Space Sci. Rev. 150, 81–124. Scholten, F. et al., 2012. GLD100: The near-global lunar 100 m raster DTM from LROC WAC stereo image data. J. Geophys. Res. 117, E00H17. Schultz, P.H., 1996. Effect of impact angle on vaporization. J. Geophys. Res. 101, 21,117–21,136. Shoemaker, E.M., Batson, R.M., Holt, H.E., Morris, E.C., Rennilson, J.J., Whitaker, E.A., 1968. Television observations from surveyor VII. In: Surveyor VII A Preliminary Report, NASA SP-173, pp. 13–81. Smith, D.E. et al., 2010. The Lunar Orbiter Laser Altimeter investigation on the lunar reconnaissance orbiter mission. Space Sci. Rev. 150, 209–241. Stopar, J.D., Hawke, B.R., Robinson, M.S., Denevi, B.W., Giguere, T.A., Koeber, S.D., 2014. Occurrence and mechanisms of impact melt emplacement at small lunar craters. Icarus, submitted for publication.

C.D. Neish et al. / Icarus 239 (2014) 105–117 Thompson, T.W., Campbell, B.A., Ghent, R.R., Hawke, B.R., 2009. Rugged crater ejecta as a guide to megaregolith thickness in the southern nearside of the Moon. Geology 37, 655–658. Ulaby, F.T., Bengal, T., East, J., Dobson, M.C., Garvin, J., Evans, D., 1988. Microwave Dielectric Spectrum of Rocks. University of Michigan Radiation Laboratory Report 23817-1-TU.

117

Wieczorek, M.A. et al., 2013. The crust of the Moon as seen by GRAIL. Science 339, 671–675. Wünnemann, K., Collins, G.S., Osinski, G.R., 2008. Numerical modelling of impact melt production in porous rocks. Earth Planet. Sci. Lett. 269, 530–539.