06)
p
p
, D
Center for Earth and Planetary Studies, Smithsonian Institution, MRC 315, Washington, DC 20560-0315, USA
b Cornell University, Space Sciences Building, Ithaca, NY 14853-6801, USA
Received 18 October 2004; revised 25 July 2005
Available online 10 November 2005
Abstract
The south polar region of the Moon contains areas permanently shadowed from solar illumination, which may provide cold traps for volatiles
such as water ice. Previous radar studies have emphasized the search for diagnostic polarization signatures of thick ice in areas close to the
pole, but near-surface regolith properties and regional geology are also important to upcoming orbital studies of the shadowed terrain. To study
regional regolith variations, we collected 70-cm wavelength, 450-m resolution, dual-circular polarization radar data for latitudes 60?90? S using
the Arecibo and Greenbank telescopes. The circular polarization ratio, ?c, is sensitive to differences in rock abundance at the surface and up to tens
of m below the surface, depending upon the regolith loss tangent. We observe significant variations in ?c, attributed to changes in the surface and
subsurface rock population, across the south polar highlands. Concentric haloes of low polarization ratio surrounding Hausen, Moretus, and other
young craters represent rock-poor ejecta layers. Values of ?c up to ?1 occur in the floors and near-rim deposits of Eratosthenian and Copernican
craters, consistent with abundant rocky ejecta and/or fractured impact melt. Enhanced ?c values also correspond to areas mapped as Orientale-
derived, plains-forming material [Wilhelms, D.E., Howard, K.A., Wilshire, H.G., 1979. USGS Map I-1162], and similar polarization properties
characterize the permanently shadowed floors of craters Faustini and Shoemaker. Small areas of very high (>1.5) circular polarization ratio occur
on shadowed and seasonally sunlit terrain, and appear to be associated with small craters. We suggest that regolith in low-lying areas near the
south pole is characterized by a significant impact melt component from Orientale, which provides a source for excavation of the block-rich ejecta
around small craters observed in this and earlier radar studies. The lower portion of the interior wall of Shackleton crater, permanently shadowed
from the sun but visible from Earth, is not significantly different in 70-cm scattering properties from diurnally/seasonally sunlit areas of craters
with similar morphology.
? 2005 Elsevier Inc. All rights reserved.
Keywords: Moon; Radar; Regoliths
1. Introduction
The lunar polar regions contain areas that are permanently
shadowed from solar illumination (Margot et al., 1999; Bussey
et al., 1999), which may provide ?cold traps? for volatiles de-
livered by comets or formed by solar wind interactions with
the regolith (e.g., Crider and Vondrak, 2000). Imaging radar,
either from the Earth during favorable librations or from an or-
bital sensor, is the only method likely to yield high-resolution
(<1 km) images of these shadowed areas. Radar also offers the
* Corresponding author. Fax: +1 202 786 2566.
E-mail address: campbellb@si.edu (B.A. Campbell).
opportunity to probe the regolith to depths of meters to tens of
meters, depending upon the radar wavelength and the regolith
loss tangent.
Previous radar studies of the south pole have emphasized
a search for cold trap locations based on interferometric topo-
graphic maps (Margot et al., 1999), and for the diagnostic
backscatter and polarization signature of thick ice deposits
(Stacy et al., 1997). These studies used radar wavelengths of
3?12.6 cm, and typically imaged areas at latitudes >80? S with
spatial resolutions of 75?125 m. Radar observations at 70-cm
wavelength permit mapping of a much larger region in a single
observing run (due to the broader antenna beam pattern), albeit
at a lower (450 m) spatial resolution. The 70-cm signals alsoIcarus 180 (20
Regolith properties in the south
70-cm radar
Bruce A. Campbell a,?
a0019-1035/$ ? see front matter ? 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.icarus.2005.08.0181?7
www.elsevier.com/locate/icarus
olar region of the Moon from
olarimetry
onald B. Campbell b
pbel2 B.A. Campbell, D.B. Cam
penetrate to greater depth, increasing our sensitivity to varia-
tions in the integrated regolith rock abundance.
The Lunar Prospector (LP) neutron spectrometer identified
enhanced hydrogen abundance near both poles, which could be
attributed to 1?2% water ice within the upper meter of the re-
golith in permanently shadowed craters (Feldman et al., 2001).
The LP data are of too coarse resolution to determine the de-
gree of localized volatile concentration. Earth-based 12.6-cm
radar observations show enhanced backscatter and circular po-
larization ratio only in association with the walls and proximal
ejecta of impact craters or likely craters smaller than can be
resolved by the imagery. No contiguous, high-backscatter re-
gions covering crater floors, similar to polar features on Mer-
cury attributed to ice (Butler et al., 1993; Harmon et al., 1994;
Black et al., 2002), are evident (Stacy et al., 1997; Campbell et
al., 2003). The Clementine spacecraft radio transmitter and an
Earth-based receiver were used to acquire bistatic radar echoes
from the Moon. These data have been interpreted to indicate
a localized concentration of ice on the lower interior wall of
Shackleton crater (Nozette et al., 1996; 2001), but their analysis
and interpretation are questioned by Simpson and Tyler (1999).
In all of these studies, the areas of permanent shadow visible
from Earth represent perhaps 1/4 of the total shadowed terrain.
The search for concentrated ice at the lunar poles thus re-
mains inconclusive. We have strong evidence for some fraction
of a hydrogen-bearing phase within the lunar regolith near the
poles, but whether this is in the form of water ice confined to
permanently shadowed craters, and to what degree such ice is
concentrated, remain uncertain. These issues will be addressed,
in part, by the thermal radiometer, ultraviolet imager, and neu-
tron spectrometer on the Lunar Reconnaissance Orbiter. The
south polar regional geology and the properties of the near-
surface regolith in shadowed terrain are of considerable im-
portance to the interpretation of these data. In this paper, we
present new dual-polarization, 70-cm wavelength radar images
that cover the entire nearside polar region (60?90? S), and 6?
7? of the farside, at a spatial resolution of ?450 m/pixel. We
use variations in backscatter strength and the circular polar-
ization ratio to study the near-surface properties of cold trap
terrain, and to trace the likely areal extent of Orientale basin
ejecta.
2. Radar data
We transmit a left-circular polarized radar signal from the
305-m Arecibo telescope, and receive lunar echoes, in both
senses of circular polarization, at the 105-m Greenbank Tele-
scope (GBT) in West Virginia. The complex-valued data, de-
noted as same-sense (SC or left-circular transmit, left-circular
receive) and opposite-sense (OC or left-circular transmit, right-
circular receive) are processed using standard delay?Doppler
techniques.
The delay?Doppler images are projected to a lunar carto-
graphic framework using ephemeris predictions that describe
the range and frequency shift of any given point with respect to
the observing station. The delay and Doppler trajectory with
time for each pixel is slightly different, with the differencel / Icarus 180 (2006) 1?7
increasing with distance from the center of the target area.
To avoid smearing of the radar image, we employ a focusing
method that maps the area in small ?chunks? after compen-
sating for the time-varying delay and frequency shifts at the
center of the chunk. Using a 3 ?s pulse and a 990 s look, we ob-
tain a horizontal spatial resolution of ?450 ? 320 m near the
pole. The final images represent averages over three separate
looks, resampled at 400-m resolution (Fig. 1). The pole is im-
aged at a radar incidence angle of ?84?. Contours of constant
incidence angle approximately parallel the radar-visible ?limb?
of the Moon.
The penetration depth of the radar signal is dependent upon
the illuminating wavelength, ?, and the loss tangent, tan ?,
of the regolith. To first order, the probing length is given by
?/(2?
?
?? tan ?), where ?? is the real dielectric constant of the
regolith. For a radar wavelength of 0.7 m and a real dielec-
tric constant of 2.7 (Carrier et al., 1991), we obtain a probing
length, in meters, of ?0.07/ tan ?. Near the poles, the high in-
cidence angle means that the refracted wave that penetrates the
surface travels at a significant angle to the vertical, so the pen-
etration depth is about 80% of the probing length. The loss
tangent of highlands material ranges from 0.01 to 0.001, though
some samples have even lower values (Carrier et al., 1991). The
70-cm penetration depth is thus 6?60 m, depending upon re-
golith dielectric properties.
Measurements in the two polarization states may be used to
determine the circular polarization ratio, ?c = SC/OC (Fig. 2).
The relative power in the two polarizations is calibrated to the
measured noise level in a portion of the data that does not con-
tain echoes from the Moon. We calculate the polarization ratio
from SC and OC images that are boxcar-filtered over 5?5 pixel
areas. This yields a 75-look average value for each location,
an effective spatial resolution of ?2 km, and an approximate
uncertainty of ?0.12 in ?c. The polarization ratio is gener-
ally higher for targets with greater diffuse scattering due to
surface or subsurface rocks. Values greater than unity are as-
sociated with coherent backscatter from thick (meter-scale or
greater) ice, or from very rugged, blocky terrain (Hapke, 1990;
Campbell et al., 1993). Values of ?c for regolith with modest
amounts (a few percent) of ice as small inclusions or thin layers
will be little different from those of normal ?dry? lunar terrain.
One aspect of our data that differs from previous lunar radar
studies is the separation of Arecibo and the GBT, which leads
to a bistatic phase angle of ?0.4? between the transmitter and
receiver. This bistatic phase angle will lead to some reduction
in the circular polarization ratio for a target characterized by
coherent backscatter (e.g., thick ice), but phase angles >1?2?
are required to entirely suppress such an effect (e.g., Fig. 1 in
Nozette et al., 2001). High circular polarization ratios in areas
with numerous surface and near-surface blocks on the scale of
the radar wavelength will not be affected.
3. Geology of the south polar regionThe new 70-cm radar data provide a synoptic view of the
south polar highlands that complements earlier geologic map-
ping from Lunar Orbiter data (Wilhelms et al., 1979). The
h poFig. 1. 70-cm wavelength, opposite-sense circular polarization radar map of the south polar region. Polar stereo projection; zero longitude at top. Latitude parallels
marked; interior circle corresponds to 85? S.
Fig. 2. 70-cm wavelength circular polarization ratio for the region shown in Fig. 1. Areas in radar shadow are set to zero. Image stretch shows polarization ratio
from black = 0.5 to white = 1.1. Concentric haloes of low polarization ratio, attributed to rock-poor ejecta, surround Zucchius, Hausen, Moretus, and Schomberger
craters.
mega-regolith within this region is comprised of contributions tures are ejecta from younger basins and craters. MaterialLunar soutfrom a number of large basin-forming impacts, including the
oldest preserved structure on the Moon, South-Pole Aitken
basin (Wilhelms, 1987). Overprinted on the ancient basin struc-lar region 3ejected from Orientale, the youngest major basin, appears to
play a very significant role in the near-surface properties of the
area surrounding the pole.
pbel
f ar
Sh = Shoemaker, Fa = Faustini, Am = Amundsen, No = Nobile, S = Shackleton, L = Idelson L, J = Wiechert J.
There are concentric regions of low ?c values surrounding
Zucchius, Hausen, Moretus, and Schomberger craters, consis-
tent with other radar-dark haloes identified by Ghent et al.
(2005) (Fig. 2). Possible haloes surround other, smaller Eratos-
thenian and Copernican craters. The haloes represent areas of
relatively rock-poor comminuted ejecta that mantles the ?av-
erage? highland regolith. The extent of these deposits is con-
siderable, and illustrates how distal crater ejecta contributes to
regolith development over time.
Enhanced (?1) ?c values are associated with the floors
of Hausen, Zucchius, Drygalski, Amundsen, and Hedervari
craters, their proximal ejecta, and a patch of highland terrain
between Drygalski and de Gerlache craters (Fig. 3). There is
also an area of enhanced ?c in a sunlit portion of Nobile
crater. The enhanced polarization ratios associated with the
floors and proximal ejecta of Eratosthenian (e.g., Hausen) and
Copernican (e.g., Zucchius) craters are readily attributed to im-
pact melt sheets with pervasive meter-scale fracturing, and to
abundant comminuted blocks formed by the impact event (e.g.,
Thompson et al., 1979).
More intriguing are the elevated polarization ratio values
associated with some ?smooth? crater floors (e.g., Shoemaker
and Faustini) and inter-crater highlands (Fig. 3). In general,
?c enhancements in sunlit areas are correlated with low-
relief, Imbrian-period plains material, mapped as ?unit Ip? by
similar deposits in Klaproth crater and between Blancanus and
Zucchius craters (Fig. 2). We do not observe enhanced ?c val-
ues in Imbrian-period basin-related materials characterized by
abundant secondary craters (units Ioc and Ioho of Wilhelms et
al., 1979), nor in Nectarian-period lineated and plains-forming
deposits (units Nbl and Ntp of Wilhelms et al., 1979). Based on
these correlations, we suggest that the floors of craters Shoe-
maker and Faustini, and local lows in the highlands surrounding
the south pole, are likely covered by smooth plains material re-
lated to Orientale.
It is interesting that the ?smooth plains? material exhibits a
higher average polarization ratio than the furrowed and pitted
basin ejecta facies. Smooth distal plains related to large basins
have been attributed to ponding of impact melt, where furrowed
and pitted terrain likely forms through local reworking of the
surface by secondary impactors (Eggleton and Schaber, 1972;
Moore et al., 1974; Wilhelms, 1987). At such great distances
from the basin, the plains-forming units may only be partially
comprised of melt rock, but this could still serve to create layers
or blocks within the regolith. These would provide subsequent
small impacts with a richer source of competent material for
fragmentation into decimeter-sized rocks than is available in
an older (e.g., pre-Nectarian or Nectarian) highland regolith,
leading to an enhanced circular polarization ratio. The same
situation is observed in the maria, where younger flows are4 B.A. Campbell, D.B. Cam
Fig. 3. Color map of circular polarization ratio, ?c , overlain on OC radar image oWilhelms et al. (1979) and attributed to basin-forming impacts.
Such material covers the floors of Amundsen (Fig. 4) and Dry-
galski (Fig. 5) craters, and enhanced polarization ratios tracel / Icarus 180 (2006) 1?7
ea near the south pole. Crater names shown as abbreviations: dG = de Gerlache,characterized by a greater surface population of rocks than
older flows with deeper regolith cover (Shorthill, 1973). A near-
surface melt-rich layer may thus explain the presence of blocky,
h poFig. 4. Portion of Lunar Orbiter Frame IV-094H1, showing the area of the south pole and Amundsen crater. Craters Shoemaker and Faustini are indicated by letters
S and F, respectively. L denotes crater Idelson L. Low-relief, Imbrian age plains (unit Ip of Wilhelms et al., 1979) in Amundsen floor, characterized by high circular
polarization ratio, marked by arrow. Surrounding pre-Nectarian and Nectarian deposits are not characterized by enhanced circular polarization ratio.
high-?c ejecta surrounding small craters in the polar shadowed
terrain (Stacy et al., 1997).
4. Observations related to possible ice deposits
On the basis of Clementine bistatic radar data, Nozette et al.
(2001) suggest that the lower interior wall of Shackleton crater
has some coverage of mixed ice/regolith deposits. The 70-cm
radar penetrates to greater depth than the 12.6-cm signals used
by earlier Arecibo or Clementine studies, and is expected to re-
veal any contiguous, meter-scale thicknesses of ice within at
least the upper 5 m of the highland regolith. In general, we do
not observe contiguous (i.e., 10-km scale), high-?c signatures
on the floors of craters such as Shoemaker and Faustini. The
highest ?c values, up to ?2, occur in patchy clusters on the
floors of both shadowed and sunlit craters (Fig. 3). Based on
Lunar Orbiter photos, higher-resolution radar data (Stacy et al.,
1997; Margot et al., 1999), and the radar scattering properties
of terrestrial rugged terrain (Campbell et al., 1993), we suggest
that these patterns reflect the proximal ejecta blankets of abun-
dant small craters.
The circular polarization ratio of crater walls varies con-
siderably with their morphology. Large, older craters with ter-
of near-surface rugged blocks. Smaller craters with sharply-
defined rims, such as Shackleton, Idelson L, and Wiechert J,
have high ?c values (?1). Craters of similar size with less dis-
tinct rims, such as de Gerlache, have little polarization ratio
enhancement. Given that high ?c values occur on both shad-
owed and diurnally/seasonally sunlit crater walls, we concur
with earlier assessments (Stacy et al., 1997) that these variations
are likely due to surface morphologic properties rather than to
near-surface ice deposits.
5. Conclusions
We present new 70-cm wavelength, dual-circular polariza-
tion radar data for the Moon?s south polar region (60?90? S),
collected using the Arecibo and Greenbank telescopes. We
observe significant variations in ?c, attributed to changes in
the surface and subsurface rock population, across the south
polar highlands. Concentric haloes of low polarization ratio
surrounding Hausen, Moretus, and other young craters rep-
resent rock-poor ejecta layers. Values of ?c up to ?1 occur
in the floors and near-rim ejecta of Eratosthenian and Coper-
nican craters, consistent with abundant rocky ejecta and/or
fractured impact melt. Enhanced ?c values also correspondLunar soutraced walls (Drygalski, Amundsen) tend to have moderate val-
ues of ?c. Younger craters such as Hausen have very strong
polarization ratio enhancements due to a greater populationlar region 5to areas mapped as Orientale-derived, plains-forming mater-
ial (Wilhelms et al., 1979), and similar polarization properties
characterize the permanently shadowed floors of craters Faus-
pbelFig. 5. Portion of Lunar Orbiter Frame IV-193H1, showing the central peaks and eastern floor and rim of Drygalski crater. The arrows mark low-relief Imbrian-age
plains units (unit Ip of Wilhelms et al., 1979) that cover the northern crater floor and are perched on the rugged rim topography; both areas are marked by high
circular polarization ratio.
tini and Shoemaker. Small areas of very high (>1.5) circular
polarization ratio occur on shadowed and sunlit terrain, and
appear to be associated with small craters. We suggest that re-
golith in low-lying areas near the south pole is characterized
by a significant impact melt component from Orientale, which
may explain the block-rich ejecta around small craters observed
in this and earlier radar studies. The lower portion of the interior
wall of Shackleton crater, permanently shadowed from the sun
but visible from Earth, is not significantly different in 70-cm
scattering properties from diurnally/seasonally sunlit areas of
craters with similar morphology.
Acknowledgments
This work was supported in part by grants from the NASA
Planetary Geology and Geophysics Program. The authors thank
J. Chandler for producing the lunar ephemeris files, J.-L. Mar-
got for assistance in setting up the radar data collection and
nell University under a cooperative agreement with the National
Science Foundation (NSF) and with support from NASA. The
Green Bank Telescope is part of the National Radio Astron-
omy Observatory, a facility of the NSF operated under cooper-
ative agreement by Associated Universities, Inc. B.R. Hawke,
D.B.J. Bussey, C.M. Pieters, M.S. Robinson, and an anonymous
reviewer provided helpful comments.
References
Black, G.J., Campbell, D.B., Harmon, J.K., 2002. New 70-cm wavelength radar
images of Mercury?s north polar region. Lunar Planet. Sci. XXXIII. Ab-
stract 1946.
Bussey, D.B.J., Spudis, P.D., Robinson, M.S., 1999. Illumination conditions at
the lunar south pole. Geophys. Res. Lett. 26, 1187?1190.
Butler, B.J., Muhleman, D.O., Slade, M.A., 1993. Mercury: Full-disk radar im-
ages and the detection and stability of ice at the north pole. J. Geophys.
Res. 98, 15003?15024.6 B.A. Campbell, D.B. Camprocessing, and the staff of Arecibo and the GBT for observ-
ing assistance. The Arecibo Observatory is part of the National
Astronomy and Ionosphere Center, which is operated by Cor-l / Icarus 180 (2006) 1?7Campbell, B.A., Arvidson, R.E., Shepard, M.K., 1993. Radar polarization prop-
erties of volcanic and playa surfaces: Applications to terrestrial remote
sensing and Venus data interpretation. J. Geophys. Res. 98, 17099?17113.
Lunar south polar region 7
Campbell, B.A., Campbell, D.B., Chandler, J.F., Hine, A.A., Nolan, M.C., Per-
illat, P.J., 2003. Radar imaging of the lunar poles. Nature 426, 137?138.
Carrier, W.D., Olhoeft, G.R., Mendell, W., 1991. Physical properties of the lu-
nar surface. In: Heiken, G. (Ed.), Lunar Sourcebook: A User?s Guide to the
Mean. Cambridge Univ. Press, New York.
Crider, D.H., Vondrak, R.R., 2000. The solar wind as a possible source of lunar
polar hydrogen deposits. J. Geophys. Res. 105, 26773?26782.
Eggleton, R.E., Schaber, G.G., 1972. Cayley formation interpreted as basin
ejecta. Apollo 16 Prelim. Sci. Report, 29-7?29-16.
Feldman, W.C., and 11 colleagues, 2001. Evidence for water ice near the lunar
poles. J. Geophys. Res. 106, 23232?23252.
Ghent, R.R., Leverington, D.W., Campbell, B.A., Hawke, B.R., Campbell,
D.B., 2005. Earth-based observations of radar-dark crater haloes on
the Moon: Implications for regolith properties. J. Geophys. Res. 110,
doi:10.1029/2005JE002425. E09002.
Hapke, B., 1990. Coherent backscatter and the radar characteristics of outer
planets satellites. Icarus 88, 407?417.
Harmon, J.K., Slade, M.A., Velez, R.A., Crespo, A., Dryer, M.J., Johuson, J.M.,
1994. Radar mapping of Mercury?s polar anomalies. Nature 369, 213?215.
Margot, J., Campbell, D.B., Jurgens, R.F., Slade, M.A., 1999. Topography of
the lunar poles from radar interferometry: A survey of cold trap locations.
Science 284, 1658?1660.
Moore, H.J., Hodges, C.A., Scott, D.H., 1974. Multiringed basins?Illustrated
by Orientale and associated features. Proc. Lunar Sci. Conf. V, 71?100.
Nozette, S., Spudis, P.D., Robinson, M.S., Bussey, D.B.J., Lichtenberg, C.,
Bonner, R., 2001. Integration of lunar polar remote-sensing datasets: Ev-
idence for ice at the lunar south pole. J. Geophys. Res. 106, 23253?
23266.
Nozette, S., Lichtenberg, C.L., Spudis, P., Bonner, R., Ort, W., Malaret, E.,
Robinson, M., Shoemaker, E.M., 1996. The Clementine bistatic radar ex-
periment. Science 274, 1495?1497.
Shorthill, R.W., 1973. Infrared atlas charts of the eclipsed Moon. Moon 7, 22?
45.
Simpson, R.A., Tyler, G.L., 1999. Reanalysis of Clementine bistatic radar data
from the lunar south pole. J. Geophys. Res. 104, 3845?3862.
Stacy, N.J.S., Campbell, D.B., Ford, P.G., 1997. Radar mapping of the lunar
poles: A search for ice deposits. Science 276, 1527?1530.
Thompson, T.W., Roberts, W.J., Hartmann, W.K., Shorthill, R.W., Zisk, S.H.,
1979. Blocky craters: Implications about the lunar megaregolith. Moon
Planets 21, 319?342.
Wilhelms, D.E., 1987. The Geologic History of the Moon. USGS Prof. Paper
1348.
Wilhelms, D.E., Howard, K.A., Wilshire, H.G., 1979. Geologic map of the
south side of the Moon. USGS Map I-1162.