JOURNAL
Surface
OF GEOPHYSICAL
RESEARCH,
Characteristics
VOL. 90, NO. B8, PAGES 6873-6885, JULY 10, 1985
of Venus
Derived
From
Pioneer
Venus
Altimetry, Roughness,and Reflectivity Measurements
JAMES
W. HEADIII, ALANR. PETERFREUND,
ANDJAMES
B. GARVINx
Departmentof GeologicalSciences,Brown University Providence,Rhode Island
STANLEY H. ZISK
Massachusetts
Institute of Technology/NorthEast Radio ObservatoryCorporation,Haystack Observatory,Westford
The three primary data setsfor the Pioneer Venus orbiter radar experiment(topography, roughness,
and reflectivity)contain important information about the geologicaland textural characteristicsof the
surfaceof Venus. We have subdividedthe range of roughnessand reflectivityvaluesinto three categories
as follows: roughness,in degreesrms slope: relatively smooth (1ø-2.5ø),transitional from smooth to
rough(2.5ø-5ø),and relativelyrough(> 5ø);and Fresnelreflectivity:surfacesdominatedby soil or porous
material(<0.1), surfacesdominatedby rock material(0.1-0.2),and surfaceswith a significantpercentage
of anomalouslyhigh dielectric material (>0.2). We have analyzed each of these data sets and their
relationshipsto each other in order to define areas of the surfacethat are characterizedby distinctive
properties(e.g., rough rocky surfaces,smooth soil surfaces).We then describethe abundanceand areal
distributionof suchareasand locally calibratethe geologicalsignificance
of someof the surfacetypesby
examininghigh-resolutionimagesfrom spacecraftand earth-basedobservatories.We find that the majority of Venus is coveredby regionallycontiguousrock and bedrock surfaces.Many of the smooth
surfaceswe interpretto be of volcanicorigin, most likely lava flows,while roughersurfacesare locally
characterizedby tectonicdeformationof severaltypes. Soil surfacescover lessthan about 27% of the
planet and are generallypatchy in their distribution. On the basisof the distribution of thesesurfaceswe
seeno evidencefor the extensivepreservationof an ancient global regolith or for widespread,topographicallycontrollederosion,lateral transport, and sedimentation.The small percentageof the surface
of Venuscharacterizedby high-dielectricmaterial appearsto originatefrom severalprocesses
including
primary lava flows probably containingenrichmentsof high-dielectricmaterials,suchas metal or metal
oxides(e.g.,Theia Mons in Beta Regio),and exposureof high-dielectricmaterialsby tectonicdeformation
(e.g.,Maxwell Montes in Ishtar Terra). Theseglobal data set correlationsprovide a fundamentalframework for understandingthe nature of the surfaceof Venus and will permit extrapolationof local and
regionalfindingsfrom futuregeochemical
and imagingexperimentsto a globalcontext.
INTRODUCTION
been revealed by Pioneer Venus (PV) observations to be di-
In this paper we subdividethe range of valuesfor roughness
and reflectivity into a small number of categories(high, intermedite, and low intervals),and we examine the distribution of
verse at scales from
each of these subdivisions
The nature
of the venusian
surface from
tens to hundreds
-65øS
to 78øN has
of kilometers.
Three
primary data sets were derived from the PV orbiter radar
(17-cm wavelength)observations:global topography, rms surface slopes(roughness),and radar reflectivity [Pettengill et al.,
1980a, b, 1982; Masursky et al., 1980]. On the basis of absolute elevation and spatial associationof topographic features,
Masursky et al. [1980] defined several global topographic
provinces.Global maps of roughnessand reflectivity showing
many subdivisionswithin the range of data have also been
used to characterize the planet's surface [Pettengill et al.,
1980b, 1982; Basilevskyet al., 1982; McGill et al., 1983]. The
roughnessand reflectivity data setshave also been evaluated
statisticallyand correlated with elevation [Garvin et al., 1983a,
b, 1984a]. Correlations of the three PV data sets have been
qualitatively described[Masursky et al., 1980; McGill et al.,
1983]. Previous efforts to characterize quantitatively these
correlations[Basilevskyet al., 1982; Schaberet al., 1982; Davis
and Schaber,1984] have involved a variety of statistical approaches incorporating various computerized clustering techniques.
• Now at GeophysicsBranch, NASA Goddard Space Flight
Center, Greenbelt,Maryland.
Copyright 1985 by the AmericanGeophysicalUnion.
Paper number 5B0197.
0148-0227/85/005B-0197505.00
6873
on the surface of Venus. The subdi-
visions provide a general concept of the nature of the surface
as follows: roughnessexpressedin degreesrms slope: lø-2.5 ø
(relatively smooth), 2.5ø-5.0ø (transitional, smooth to rough),
and >5.0 ø (relatively rough); and Fresnel reflectivity: <0.1
(surfaceswith a majority of porous material), 0.1-0.2 (surfaces
with a majority of material comparable to terrestrial rock),
and >0.2 (surfaceswith a significant percentage of anomalously high-dielectricmaterial). We then assessthe degree of
correspondenceof the subdivisionsof roughnessand reflectivity with the physiographic/topographicprovinces defined
by Masursky et al. [1980]. Finally, we assessthe degree of
correspondencebetweenthe roughnessand reflectivity subdivisionsby mapping globally the distribution of combinations
of the subdivisionsof thesetwo parameters(e.g.,mapping the
distribution of regions characterized by high values of both
roughness and reflectivity). The relationship between radar
roughnessand reflectivity and the relationship of these two
parameters to topography are of interest due to (1) possible
variations in crustal structure and composition which may be
revealedby variationsin roughnessand reflectivityand which
may vary as a function of elevation,(2) geochemicalprocesses
that may be pressure-temperature
dependentover the range of
surfacepressures(60 bars) and temperatures(100 K) on Venus
[Florenskyet al., 1978; Nozette and Lewis, 1982], and (3) sedimentation and weathering that may be a function of absolute
6874
HEAD ET AL.: SURFACECHARACTERISTICS
OF VENUS
elevation and/or slopesand which may be reflectedin terms of
variations in surface roughenss or proportions of
soil/rock/high-dielectricmaterial.
METHOD
The three PV data sets used to produce the maps for this
study include data collected and processedas of December
1982. Analyseswere done utilizing the PV data mapped into a
Mercator projection with a spatial resolution of løx 1ø
(equivalent to 105 km x 105 km at the equator and 53
km x 105 km at +60ø). The maps were produced using a
5ø x 5ø boxcar filter with uniformly weighted coefficientsin
order to fill in those 1ø x 1ø cells for which valid PV data were
unavailable. Approximately 10% of the cells required filling.
Surface areas were estimated from these Mercator maps by
using a correction factor for latitude. Banding observed in
these maps is due primarily to absent or invalid data associated with a specific PV spacecraft revolution (e.g., orbital
ground tracks) and, to a lesserextent, with orbit-to-orbit variations. Geographic place names for Venus used for reference
in the following discussioncan be found on the U.S. Geological Survey (USGS) 1'50,000,000 maps [USGS, 1981, 1984]
and in the works by Strobell and Masursky [1983] and Masurskyet al. [1984] and are shown in Plate 1.
Global subdivision of topography into provinceshas been
previouslyproposedby Masursky et al. [1980] as follows' (1)
lowlands were defined as regions with altitude <6051.0 km
radius (i.e., <0 km elevation), (2) rolling plains as regions
between 6051.0 and 6053.0 km (0.0-2.0 km), (3) highlands
(>6053.0 km). We further subdivide the highlands of Masursky et al. [1980] into highlands(regionsbetween6053.0 and
6055.5 km; 2.0-4.5 km), and mountainous regions (>6055.5
km' >4.5 km) (Plate 2). The mean planetary radius of 6051.2
km [Pettengill et al., 1980b] (updated to 6051.9 km for the
December 1982 data set by Garyin et al. [1984a]) is within the
rolling plains unit, which covers •75% of the planet. The
subdivisionsof Masursky et al. [1980], amended by us, define
topographic provinceson Venus that are spatially distinct and
serveto outline specificgeographicregions.
Subdivisions for radar roughness and reflectivity were
chosen on the basis of standard interpretations of radar
roughness and reflectivity measurements(reviewed by Pettengill et al. [1980b, 1982] and Garyin et al. [1983a, b, 1984a]
and summarized here). The rms surface slope is a measure of
the small-scale(0.5-100 m) roughnessaveragedover the radar
resolution element. The rms slope is derived from a model
based on the Hagfors scattering law for the quasi-specular
radar return from a planetary surface as follows [Ha•;lfors,
1970]'
ao(O)= (PoC/2) (cos'*0 + C sin2 0)-3/2
(1)
where ao is the radar cross section per unit surface area at
anglesof incidence0, Po is the Fresnel reflection coefficientat
normal incidence angle, and C is the Hagfors parameter. In
Hagfors' original model calculations [Haqfors, 1964], the rms
slope of the reflecting (specular)surfacefacets())wavelength)
measurementis approximately0.5 m to tens of meters[Pettengill et al., 1980b]. In comparison with similar radar
measurements
for the moon and Mars (seereviewsby Pettengill [1978] and Ostro [1983]), Venus appearsto be relatively smooth.Three subdivisions
in rms slopewere chosen:
(1) smooth,1ø-2.5ø, typicalof the smoothestregionsof Mars,
(2) transitionalfrom smoothto rough,2.5ø-5.0ø, typicalof the
lunar maria, and (3) relativelyrough, > 5.0ø, typicalof lunar
highlandsand the roughestsurfaceson Mars [Simpsonet al.,
1984] (Plate 3). The transitionalrange comprises-,•46% of
the observedsurfacearea of the planet. It could either be
inhomogeneous,
possiblycontaininga mixtureof both smooth
and rough elements,or could representa distinctsurfacemorphology.
Reflectivity
valuesare derivedfromthe scattering
modelby
fitting the FresnelreflectioncoefficientPoto the data [Pettengill et al., 1982]. The reflectivityis a function of the complex dielectricconstante, where
/90 =
+
(2)
The entire Hagfors theory [Hagfors, 1964], including this
equation, concernsonly the quasi-specularsurfacecomponent
and none of the diffuse(random scattering)component.It is
possible,therefore,that if a surfaceis coveredby a large fraction of random-scatteringelements,the reflectivity calculated
from the remaining quasi-specularecho will be less than the
Fresnel reflectivity of the surfacematerials. The complex dielectric constant
is a characteristic
of the surface material
and
includes a dependenceon the volume conductivity and the
bulk density (porosity) as well as rock chemistry. Krotikov
[1962] and Krotikov and Troitsky [1963] have measureddielectricpropertiesof a variety of dry terrestrialrocks ranging
in densityfrom pumice,p = 1000-1800kg m-3, to dunites,
p- 3300 kg m-3 (includingbasalts,glasses,and granites).
They found an approximate relationship between density and
dielectricconstant. For dry, nonconductingterrestrial materials, the bulk density d (in kilograms per cubic meter) can be
approximated by
d •
x/•-I
0.5
X 103
(3)
where the constant 0.5 was empirically derived for measurementsat PV radar wavelengths[Krotikov and Troitsky, 1963].
Campbelland Ulrichs [1969] also measuredthe dielectricconstant and loss tangent for a variety of geologicmaterials, including both solid rocks and powdered samplesof identical
rocks.At leastfor the solid rock samplesit is observedthat (3)
remainsa good approximation of bulk density [Garvin et al.,
1985]. Hence for terrestrial materials the observed dielectric
constant
e for solid rock exhibits
a lower limit
of about 4
(reflectivityof 0.11). Observedvaluesof reflectivitylower than
this are likely to be due to a significantfraction of porous
materials (e.g, soils),with the expectedfraction of suchmaterial increasingas the observedreflectivity decreases.For examwas found to be equal to 180/7rC
•/2 for low to moderate ple, the lunar surfaceis dominatedby a porousregolithwith a
roughness (C))100). As the calculations are based on a mean reflectivityof 0.07 [Tyler and Howard, 1973].
model, however, the resulting valuesfor rms slope are only an
For Venusthe observedmean reflectivityis 0.13,which sugindication of the angular distribution of scatteringobjectson geststhat either the fraction of porous material is much lower
the surface.The larger the rms slope, the greater the amount on the Venusian surface than on the moon or that the domiof surface undulation
or surface block cover. The mean rms
nant materials have a remarkably higher bulk dielectricconslope for Venus is 2.65ø+ 0.75ø [Pettengill et al., 1980a; stant [Pettengill et al., 1982]. In light of the observedrange in
Garyin et al., 1984a] where the scale length for the roughness reflectivity over the planet's surface,we have subdividedthe
HEAD ET AL.: SURFACE CHARACTERISTICSOF VENUS
data set into three parts: (1) <0.1, to be less than any measured terrestrial solid rock, low dielectric constant (e < 4),
almost certainly incorporating a significantfraction of porous
material, (2) 0.1-0.2, moderate dielectric constant, probably
comprising a large fraction of consolidated material if of terrestrial or lunar mineralogy [01hoeft and Strangway, 1975],
and (3) >0.2, at or beyond the upper end of the range of
terrestrial rocks, high dielectric constant (e > 9), indicating
mainly consolidatedmaterial with perhaps some anomalously
high-dielectric material representing mineralogies probably
enriched in FeS2, Fe, and/or Ti oxides(Plate 4). The moderate
range, which comprises 67% of the observed surface area of
the planet, is likely to include a significantproportion of rocks
or bedrock with only a minor contribution from porous
("soil") fractions. Even soil derived from high-dielectric minerals should make only a minor contribution, since Campbell
and Ulrichs [1969] determined that the dielectric constants of
powdered rock samplesare universally low (e.g., e < 4), independent of the properties of the original rock. As has been
noted elsewhere,this surface type, the largest fraction of the
Venusian surface,is unlike the unconsolidated soil typical of
the moon and much of Mars.
The map distribution of correspondencesbetween roughness and elevation subdivisions is shown in Plate 5, between
reflectivity and elevation subdivisionsin Plate 6, and between
roughnessand reflectivity subdivisionsin Plate 7. Note that
the three parameters (roughness,reflectivity, and elevation) are
6875
toward Ishtar Terra show low roughnessvalues.Atalanta Planitia, in contrast, is dominated by widespreadregions of low
roughness.
The rolling plains (0.0-2.0 km) display broad regions of
both smooth and transitionalroughness.Transitional roughnessareas often flank highland regions,most notably around
Beta-Phoebe Regio, but also in narrower, less continuous
zones around Ishtar and Aphrodite Terra. Broad circular features, such as the Nightingale-Earhart region of eastern
Tethus Regio, show intermediate roughnessvalues. Many of
the small elevated plateaus (e.g., Tellus, Alpha Regio) contained within this topographic province have transitional to
high roughnessvalues. Rough units occur as either small regional clusters or isolated features. Chasmata contained
within Aphrodite Terra, while poorly resolvedat this spatial
resolution,appear to be relatively smooth.
Most highlands(2.0-4.5 km) appear transitionalin roughness. Smooth regions can be observed in Lakshmi Planum,
northwest and west of Maxwell
Montes, west of Akna
Montes, and in isolatedsectionsof Aphrodite. Rough regions
occur adjacent to the mountainous terrain within the highlands, as seen,for example,in the western and central highlands of Aphrodite and eastern Ishtar Terra. Mountainous
regions(> 4.5 km) are mostlytransitionalor rough,with only
isolated occurrences of smoother
surfaces.
Elevation VersusReflectivity
Spatial correspondencesof radar reflectivity and elevation
are shownin Plate 6 (also comparePlates 2 and 4). Elevations
are representedby huesand reflectivitysubdivisionsby intensity variations.A lessdistinctrelationshipexistsbetwenelevation and reflectivity than that observedbetweenelevation and
roughnessfrom a statisticalstandpoint [Garvin et al., 1984a]
as well as from generalmap unit trends.While a generaltrend
of increasingreflectivitywith elevationis apparent,for a given
elevationrange a wide range of reflectivitymap units are observed(Plate 6).
transitional at their boundaries.
The majority of lowlands are moderatein reflectivity,sugRESULTS
gestinga predominantlyrock surface.Lower-reflectivityunits
dominate Lavinia Planitia and occur as distinctivepatchesin
Elevation Versus Roughness
the otherwise intermediate-reflectivityGuinevere and Sedna
Spatial correspondences
of three divisionsin roughnessand planitiae. The most distinctiveoccurrencesof high-reflectivity
four divisionsin elevation are shownin Plate 5. Using hue to patches are in Atalanta Planitia. On the basis of these obserdistinguishdivisions in elevation and intensity to distinguish vations it is clear that there is a diversity of reflectivityunits
divisionsin roughness,a pattern of increasingroughnesswith
both within and between planitia (e.g., within Sedna and beelevation is apparent (also comparePlates 2 and 3). As noted tween Sednaand Atalanta planitiae).
in previous studies [Pettengill et al., 1980b; Masursky et al.,
Rolling plains are also mostly moderate in reflectivity,
1980; McGill et al., 1983; Garyin et al., 1983a], highestregions implying a predominance of surface materials with dielectric
are generally roughest, and lowest regions are relatively properties similar to terrestrial rocks. Low-reflectivity units
smooth,although not necessarilythe smoothestregionsfound tend to be adjacent to highland regions, such as Beta and
on Venus. In detail, however,noteworthy deviationsfrom this Aphrodite Terra, but are not evenly distributed around the
trend are observed.
entire highland perimeters.Several regiones are also domiMost lowlands (<0.0 km elevation) are smooth or transi- nated by low reflectivity(Alpha, Tellus) as is Lada Terra, the
tional from smoothto rough. Within the lowlandsthe regions rolling plains area at high southern latitudes. Chasmata, in
of low and moderateroughness(darker hues)are spatially well general,appearto containmaterial with low reflectivity.Highdefined (i.e., occur as clustersof løx 1ø cells).In contrast, reflectivity units are widely distributed but occur mostly in
high-roughnessunits in the lowlands occur in isolated patches isolated small areas.A notable exceptionto this is the region
(i.e., single løx 1ø cells of high roughnesssurrounded by west of Atalanta Planitia (the Nightingale-Earhart region of
smoother surfaces).Small topographicdepressions(<10 • easternTethus Regio).
km2) tend to havemoderateroughness
values,while broader
Highland regionsgenerallyshow a pattern of highestrefleclowlands are both smooth and transitional. The pattern of tivity adjacentto mountainousterrain, with decreasingreflecroughnessin the broader lowlandsshowsregional clustering. tivity away from thesepeaks.This pattern is well illustrated in
In Sedna and Guinevere planitiae, for example, areas toward the western and central highlands of Aphrodite and in northBeta Regio show intermediate roughnessvalues, while those ern Beta Regio. In eastern Ishtar Terra the pattern of refleccontinuousover the entire data range. Therefore, although the
resulting surface units are defined within a specificrange of
values (for example,regionsof low roughnessand intermediate reflectivity have values of 1ø-2.5ø rms slope and 0.1-0.2
reflectivity, indicating that the terrain is likely to be smooth
and composedof a large fraction of rock material), they are in
fact gradational with each other becauseroughness,reflectivity, and elevation representa spectrumof values.Thus, as is
common with terrestrial geologicunit definition, the units are
6876
HEAD ET AL.' SURFACECHARACTERISTICSOF VENUS
TABLE 1. Summaryof Roughness-Reflectivity
Subdivisions
Subdivision
Surface
Area,
rms Slope,
Reflectivity %
deg
A
1.0ø-2.5 ø
0.02-0.1
11.5
B
2.50-5.0 ø
0.02-0.1
13.5
C
5.0ø-10.0 ø
0.02-0.1
2.2
D
1.0ø-2.5 ø
0.1-0.2
36.5
E
2.50-5.0 ø
0.1-0.2
28.6
F
5.0ø-10.0 ø
0.1-0.2
2.3
G
1.0ø-2.5 ø
0.2-0.5
2.1
H
2.50-5.0 ø
0.2-0.5
2.3
I
5.0ø-10.0 ø
0.2-0.5
1.0
Type Region
(N Latitude,
Longitude)
Lavinia Planitia
(-40 ø, 340ø)
Alpha Regio
(- •5 ø, 5ø)
Tellus Regio
( + 35ø, 80ø)
NW of W Aphrodite
Terra ( + 30ø, 50ø)
Beta Regio
( + 30ø, 285ø)
Rhea Mons
( + 33ø, 282ø)
W Atalanta Planitia
( +60 ø, 150ø)
SW Atalanta Planitia
( + 55ø, 135ø)
Maxwell Montes
( + 63ø, 4ø)
tivity units follows a more regional distribution than the pattern observed around Beta and Aphrodite Terra. Much of
easternIshtar is dominated by low reflectivityvaluesin contrast to the dominance
of intermediate
values in the Lakshmi
Planum region of westernIshtar Terra.
Mountainous regions,as noted by Pettengill et al. [1982],
contain most of the highest-reflectivitymaterial observedon
Venus. These high-reflectivity values are in excessof what
would be expectedbased on the bulk density-reflectivityrelationship (equation (3)) and as such are inferred to indicate
the presenceof a considerablepercentageof high-dielectric
material. The PV radiometric observations of exceptionally
low emissivityat theselocations support the presenceof such
high-dielectricmaterials [Ford and Pettengill, 1983]. Mountainous regions showing these high values include Maxwell,
some peaks within Akna and Freyja Montes, Theia Mons,
and the highestterrain within Aphrodite.Not all mountainous
terrains, however,are characterizedby high reflectivityvalues;
for exampleRhea Mons and most of Akna and Freyja Montes
have surfacematerial with moderate reflectivity.
Interpreted
Characteristics
smooth; soils d- rock
transitional;soilsd- rock
rough' soilsd- rock
smooth; rock d- soil
transitional' rock d- soil
rough' rock d- soil
smooth; high dielectrics-/-_rock
transitional' high dielectricsd- rock
rough; high dielectricsd- rock
isolatedcells.Significantexamplesof this unit are locatedin
Lavinia Planitia and the rollingplainsto the east(Lada Terra)
as well as adjacentto mountainswithin Ishtar Terra and
alongthe flanksof easternIshtar Terra.
Unit B (green)containssurfacesthat are transitionalin
roughness
and havea low reflectivity.We referto this as the
intermediate-roughness
soil unit. This unit is proportionally
lessprevalentin lowlandsand more commonin highlands.
These surfacesappear closelyassociatedwith the edgesof
highlands
(Aphrodite,Beta),but the unitsdo not fullycircumscribethe highlands.Many of the smallelevatedplateaus(regiones)are characterized
by thisunit. The mostpredominant
occurrences
of this unit in the highlandsare in easternIshtar
and Akna and Freyja Montes.
Unit C (dark green)is the third of the low-reflectivitysurfacesand is characterizedby high radar roughness.
We refer
to this as the roughsoil unit. The unit occurspredominantly
in areasadjacentto and within highlands.Occurrences
within
lowlands are isolated. Most of the chasmata are contained
within this unit. Other occurrences of the unit include Tellus
percent of the surfaceis containedwithin four of the nine
definedroughness-reflectivity
units. Two low-reflectivityunits,
and Phoeberegionesand the highlandsin easternIshtar.
Unit D (light blue)is characterized
as smoothand of moderate reflectivityand is the most widespreadsurfaceunit
(Table 1). It is probablydominatedby surfacematerialswith
rocklike dielectricproperties,and we thereforerefer to it as
the smoothrock unit. This unit occurspredominantlyin the
lowlandsand rolling plainswith only isolatedoccurrences
in
A and B, are smooth and transitional in roughness,respec-
thehighlands
andis virtuallynonexistent
in themountains.
In
tively, and accountfor • 25% of the observedVenusiansurface. Two moderate-reflectivity units, D and E, are also
smooth and transitional and make up • 65% of the observed
surface.Table 1 presentsa summary of the unit parameters,
the lowlandsand rolling plainsthe unit usuallyoccursin con-
RoughnessVersusReflectivity
Roughness-reflectivity
comparisonsare shown in Plate 7
with nine subdivisions,increasingin reflectivityalong the horizontal axis and in roughnessalong the vertical axis. Ninety
their distribution, and characteristicsof the individual units
discussed below.
tiguouspatches,
a majorpatchoccurring
in AtalantaPlanitia.
A very largeirregularregionof this unit is seenat southern
mid- and high latitudesextendingfrom about 120ø to 330ø
longitude.A distinctivelinearpatchis seenextendingfrom
just southof Maxwell Monteseastwardto TethusRegio,a
Unit A (light green)containssurfacesthat are smoothand distanceof over 3000 km. In the highlands,notable examples
have a low reflectivity,implying that a majority of the surface of occurrence include Lakshmi Planum and Hathor Mons.
Unit E (sky blue) is characterized
by moderateroughness
is likely to be coveredby porousand unconsolidated
material.
Thus we refer to unit A as the smooth soil unit. This unit is
andreflectivityvalues,andwereferto thisastheintermediateroughness
rock unit. It occursin the sameproportionat all
slightly more prevalent proportionally (with respectto areal
divisionsof global topography)in lowlandsand rolling plains
than in highlandsand mountains.The occurrenceof the unit
tial portionof GuineverePlanitiaand asisolatedcellsin most
in all topographic
regionsis as smallpatches
(< 104km2) or
other lowland areas.With'?. the rolling plains, two broad re-
elevation levels. In the lowlands the unit occurs as a substan-
HEAD ET AL.: SURFACECHARACTERISTICSOF VENUS
gions fall within the intermediate-roughnessrock unit. The
first, which occurs to the east of Beta Regio, is particularly
notable in that it contains all of the Venera landing sites for
which images are available [Florensky et al., 1977, 1983].
Radar characterisicsfor the areas surrounding these specific
sites have been derived from PV observations[Garvin et al.,
1984b] and are consistentwith values contained within this
unit (roughness= 30-4ø and reflectivity=0.10-0.15). The
other major occurrencesof this unit are in regions to the
north and east of Atla Regio. Within the highlandsthe unit is
common in Metis Regio, Vesta Rupes,and Beta Regio and in
much of Aphrodite. Isolated occurrencesof the unit are found
in most mountainousregions.
Unit F (dark blue) contains rough surfaces of moderate
reflectivity, and we refer to it as the rough rock unit. Only
isolated
occurrences
of the unit
are found
in lowlands
and
rolling plains. Within the highlandsthe unit is observednear
mountainsin Aphrodite, north of Maxwell Montes, and in the
southernpart of Tethus Regio. Significantoccurrencesof the
unit within mountainous terrain include Rhea Mons, Akna
and Freyja Montes, and a band east of Maxwell Montes.
Unit G (yellow) is characterized by smooth surfacesand
high-reflectivitymaterial. As seenin Plate 4, reflectivity values
are not as high as observedin the mountains (unit I) but are
suchthat a significantpercentageof high-dielectricmaterial is
likely to be present. The most distinctive occurrence of this
unit is in a region east of Tethus Regio surroundingEarhart
and Nightingale and extending into Atalanta Planitia. Other
occurrencesare more localized with significantclusterssouthwest of Akna Montes and as patchesin the widespreadunit D
at southernmid- to high latitudes.
Unit H (orange) is transitional in roughnessand contains
materials with high radar reflectivity.This unit is proportionally more common in highlands and mountainous regions.
Occurrenceswithin lowlandsand rolling plains are limited to
small patches and isolated occurrences,with well-defined examples in the Atalanta region, southeastof Aphrodite Terra,
and west of Alpha Regio. In the highlandsthis unit occursin
small patchesadjacent to summit areas. Examples within the
mountainsinclude Rhea Mons and part of Maxwell Montes.
Unit I (red) occurspredominantly in highlands and mountainous regionsand contains surfacesthat are rough and have
a high reflectivity. Isolated occurrencesof this unit are found,
however,in lowlands and rolling plains. Maxwell Montes in
Ishtar Terra, Theia Mons in Beta Regio, and Ovda, Thetis,
and Atla regiones within Aphrodite are the most extensive
examplesof this unit.
DISCUSSION
The analysisand subdivisionof roughnessand reflectivity
data have permitted us to plot the distribution of thesesubdivisions on the surfaceof Venus (Plates 3 and 4), to examine
the altitude distribution of thesesubdivisions(Plates 5 and 6),
and to assessthe distributionof units definedby combinations
of roughnessand reflectivitysubdivisions(Plate 7). The basis
for the subdivisionsof roughnessand reflectivity data, the
transitional nature of the boundaries of these subdivisions,
and thus the nature of the units shown in Plate 7 have been
discussedin previous sections.The units resulting from the
combination of thesesubdivisionsdo not uniquely define distinctive geologicalprocessesbecausethey map (1) variations
in surfaceroughnessand (2) propertieswhich can be interpreted in terms of relative proportions of soil, rock, and highdielectricmaterials on the surface.Obviously, for example,a
6877
rock surface can be produced by a wide range of geologic
processesincluding volcanism, tectonism, erosion, and sedimentation. The subdivisionsand units do provide, however,
an important characterization of the surface properties of
Venus which permits an initial assessment
of classesof geologic processesthat might be operating to form and modify the
surfaceof Venus. In this sectionwe first examineimaging data
which allow us to identify the geologicalprocessesoccurring
within some regions of our mapped units. We then review the
distribution of these subdivisionsand units and develop a
seriesof interpretations and predictionsconcerningthe nature
of the surface of Venus.
Moderate-reflectivity units contain surface materials with
dielectric constantsconsistentwith terrestrial rocks. The paucity of a soil cover could be indicative of extensive bedrock
exposuresor a high percentageof large fragmental rocks. For
example, the Venera lander sites for which images are available all exhibit blocky or bedrock surfaces.The Venera 9 area
contains a mixture of blocks and soil, while at Venera 10 the
surface is characterizedby a bedrock pavement overlain by
patchesof soil [Florenskyet al, 1977]. The region surrounding
these two sites is characterized by average reflectivity values
lying at the lower end of the intermediate (predominantly
rock) range [Garvin et al., 1984b]. The region surrounding the
Venera 13 and 14 sitesis characterizedby average reflectivity
values which lie in the middle of the intermediate (predominantly rock) range. Images of these sites show bedrock pavement at both locations, with bedrock comprising essentially
100% of the surfaceviewedby the Venera 14 spacecraft[Florenskyet al., 1983; Gart)inet al., 1984c].
Regional bedrock units could represent volcanic flow surfaces,consolidatedsedimentsof various origins, or eroded and
exposed rock of various origins; the radar reflectivity data
alone do not distinguish between these possibilities. Earthbased radar data [Campbell and Burns, 1980; Campbell et al.,
1983, 1984a] provide information on possibleorigins in some
areas. For example, the region in the rolling plains southeast
of Lakshmi Planum is dominated by a sequenceof lava flows
[Campbell et al., 1984b-I covering several hundred thousand
square kilometers. This particular region is characterizedprimarily by intermediate reflectivity units of low to intermediate
roughness(units D and E). High-resolution radar images of
Beta Regio [Campbell et al., 1984] confirm the presenceof
large volcanic structures and extensive flowlike deposits associated
with
the Devana
Chasma
rift zone. These
volcanic
depositslie predominantlyin the moderate reflectivity (rock)
units (E and F). In addition, recent orbital high-resolution
images from Venera 15/16 show that Lakshmi Planum, a
broad region of intermediate reflectivity in Ishtar Terra, is
characterized by two major calderas and widespread lava
flows [Barsukov et al., 1984]. Thus we conclude that volcanic
processescharacterizeportions of the intermediate reflectivity
(rock) units (D, E, and F) and may account for some of the
larger expansesof this unit in the lowlands and rolling plains
suchas Guinevere,Sedna,Helen, and Atalanta planitiae.
Several areas characterized by moderate reflectivity (rock)
units are interpreted to be dominated by tectonic activity,
including the banded terrain in Akna and Freyja Montes of
Ishtar Terra [Campbell et al., 1983], the region between Ut
and Vesta Rupes, and portions of the Devana Chasma rift
zone in Beta Regio [Campbell et al., 1984a, hi. These areas
tend to be characterizedby intermediate to high roughness.
Deformational processes(tectonics)appear to be responsible
at least in part for the local developmentof the intermediate-
6878
HEAD ET AL.: SURFACECHARACTERISTICSOF VENUS
areas of regional soil development, such as eastern Ishtar
Terra or Lada Terra, may be areasof tectonicactivity.
Eolian processesrepresent an additional possiblemechanism of soil transport on Venus [White, 1981; Greeleyet al.,
1984]. Evidencefor local erosion of bedrock and collection of
soil in intervening low areas is seen in the Venera lander
panoramas!-Florenskyet al., 1977], althoughthe role of eolian
processesat this scale has not been firmly established.The
possibilityexiststhat someof theseregionalsoil units may be
eolian deposits.Under the presentenvironmentalconditions,
surfacetemperatureis relatively constant,surfacewinds have
exact areal extent of soil on the surface of Venus is difficult to
relatively low velocity, and there is a lack of seasonalor latiascertain, surface units consistent with a soil interpretation tudinal variations in environmentthat might enhanceeolian
(units A, B, and C) cover only about 27% of Venus. The processes.
Thereforeit is difficult to predict the global patterns
distribution of theseunits can be an aid in limiting the various of units that might result from eolian activity, although altimodels of soil formation. The two smoothestlow-reflectivity tude variations may be important becauseof the variation in
units (Plates 4 and 7) comprise less than 25% of the surface, temperatureand pressure.Higher-resolutiondata are required
are usually spatially contiguous,occur at all elevations,and in order to identify potential eolian deposits and to assess
often occur in small patches.On the moon a global impact- their relation to these soil units.
Pyroclasticactivity offersanother mechanismfor the formageneratedregolith dominatesthe surfaceto depthsin excessof
several meters. The regional clusters of soil units on Venus tion of soil units. Extensivepyroclasticdepositsare considered
argue against an origin related to a global impact-generated unlikely in the presentenvironmentbecauseof the influenceof
ancient regolith, although remnantsof ancientregolith cannot high atmospheric pressures!-Garyin et al., 1982; Head and
be ruled out. If soil formation and accumulation(not necessar- Wilson, 1982]. In addition, soil units are not directly associily by impact comminution) scales simply as a function of ated with the distinctivevolcanic depositsand constructsrectime, then areasof extensivesoil development,suchas eastern ognized on Venus thus far [Campbell et al., 1984a, b; Head et
Ishtar Terra, Lada Terra, and Tellus Regio, may be some of al., 1985]. However, a contribution to the distribution of soil
units by pyroclasticactivity cannot be ruled out.
the oldestterrain on the planet.
In the southernhigh latitudesa vast regionis dominatedby
Soil units might also represent processesof weathering,
degradation, and lateral downslope movment of sediment, the soil units. Topographically,this soil-rich terrain is associdriven by gravity. On earth these processes,assistedby the ated with a broad elevated region (Lada Terra) within the
atmosphereand hydrosphere,result in regional sedimentand rolling plains that extendsto the southernedgeof the Pioneer
soil units in low-lying areassuchas abyssalplains,continental Venus coverage.Soil units are also common in some regions
interiors, and annuli surrounding elevated regions. A first- (Tellus,Alpha, easternIshtar) while uncommonin others(Beta
order observation for Venus is that soil units (units A, B, and Regio),suggestingpossibledifferencesin origin or surfacesof
C) are only slightly more common proportionally in the low- differingage.
lands and rolling plains than in the highlands.Comparisonof
The high-reflectivityunits indicatethe presenceof a signifiPlates 2, 4, 6, and 7 shows that soil units do not form extencant amount of high-dielectricmaterial near the surface.On
sive, continuous parts of the lowlands (for example, the floor earth, common high-dielectricmaterials include liquid water,
of Sedna,Guinevere,or Atalanta planitiae).We thus conclude freemetal, and certain minerals(for example,futile, magnetite,
that large-scalelateral movement of sedimentand soil in this ilmenite, hematite,spinel,pyrite, pyrolusite).The geologicenmode is not a dominant processin shapingthe observedsur- vironments in which high-dielectricmaterials are found on
face of Venus. Sharptonand Head [1984], in an analysisof earth are quite variable. Liquid water is ubiquitous,free metal
regional slopeson Venus and earth, find that Venus has a is extremelyrare, and high-dielectricminerals are common in
deficiencyof the very low slopestypical of continentalplat- igneous and metamorphic rocks in relatively low conforms and abyssalplains on earth.
centrations.For example,the highestconcentrationof TiO: in
There is evidence,however,that soil units occuradjacentto typical terrestrial lava flows is in the 2-3 wt % range and
some highland regions on their flanking slopes,or adjacent occursin continentalrift zones[Carmichael,1982]. Although
upland rolling plains, particularly in northern Beta Regio, direct measurements have not been made, these consouthwestof Ishtar Terra, and adjacentto portions of Aphro- centrationsare likely to be in the radar reflectivityrange of
dite Terra (Plates 4 and 6). This suggeststhat at least some intermediate,rather than high, values.Concentrationsof iron
processesof weathering and downslope movement may be and titanium oxidesin lunar lava flows are often considerably
taking place adjacentto highland regions,althoughthe asym- higher (10-12%). Olhoeft and Strangway [1975] have shown
metric development of soil units in these areas hints at com- that Apollo 11 high-titanium basaltshave dielectricconstants
plexity. Soil formation might also preferentiallyoccur in areas in the range of 10-12, which would produce radar reflectivity
of steep slopes and tectonic disruption due to fracturing, valuesof 0.25-0.30, within the high-reflectivityunit mappedin
gravity-induced movement and abrasion, or simple prefer- this study. Therefore high concentration of these highential accumulation in topographic traps. For example, on dielectric minerals occurring in primary volcanic rocks is a
earth, mountainsor ridgesformed by tectonicactivity in both possiblegeologic environment for high-reflectivityunits on
extensional (Basin and Range) and compressional(Appala- Venus. It is recognized that many of these minerals may
chian Valley and Ridge) environmentsare characterizedby changetheir dielectricpropertieswith increasedtemperature
intervening areas of sediment fill. On Venus, soil units are and pressure[Parkhomenko,1967]. Detailed data on mineral
associated with the tectonic banded terrain in Ishtar Terra
stability and dielectric properties under Venusian conditions
[Campbell et al., 1983] and with the floor of Artemis Chasma are required before specificcandidatescan be identified.Eroin Aphrodite Terra lehmann and Head, 1983]. Thus other sion and tectonic activity could expose subsurfacehighreflectivity units either by local surfacedisruption or the generation of regional topographic slopes.
The low-reflectivityunits on Venus most likely contain surfaceswith a majority of porous and unconsolidatedfine materials (soil). The rougher low-reflectivity units may contain a
sufficientfraction of diffuse-scattering
elementssuchthat their
reflectivity values have been artificially lowered by 10-15%
[Pettengill et al., 1982; Ford and Pettengill, 1984]. Possible
origins of the soil include impact-generatedregolith, volcanic
pyroclastics,in situ weathering of bedrock, and depositional
products of lateral masstransport of sediments.Although the
HEAD ET AL..' SURFACE CHARACTERISTICS OF VENUS
reflectivity materials of igneousor metamorphic origin. Other
possibleorigins for high-reflectivityunits on Venus include
elevation-dependent(temperature-pressure)
chemicalor physical weathering [Nozette and Lewis, 1982; McGill et al., 1983],
which might produce or concentratehigh-reflectivitymaterials, and other various erosional mechanisms, such as eolian
activity, which might concentratematerials on the basis of
their density.
In an analysis of Venus global surface radar reflectivity,
Pettengill et al. [1982] pointed out that highly conducting
metallic sulfides display high-dielectric values but would be
unstable to atmospheric exposure according to the calculations of Nozette and Lewis [1982]. Pettengill et al. [1982]
proposed that the regions of high reflectivity on Venus are
rocks containinga significantamount of conductingmaterials
as inclusions, favoring FeS2 (pyrite) as the inclusion. They
suggestthat pyrite may be widespreadin original crustal rock,
but that it lies in radar view only at higher elevationswhere
new surfacesare constantlybeing exposedby mechanismsof
chemical
erosion
and
lateral
sediment
movement
into
sur-
rounding lows, as envisionedby Nozette and Lewis [1982].
High-reflectivity materials are concentrated in five major
areas: central Beta Regio, Maxwell Montes in Ishtar Terra,
the Tethus Regio-eastern Atalanta Planitia region, Ovda,
Thetis, and Atla regiones in eastern Aphrodite Terra, and
6879
rocks, perhaps they occur only in the freshestvolcanic deposits and in materials exposedby recent tectonic/erosional
activity (e.g., Maxwell Montes). Again, the observed high
roughnessvaluesfor theseareasare consistentwith this interpretation. Following this sequenceof logic, one might favor a
model in which high-dielectricmaterial is exposedduring the
emplacementof lava flows but is diluted and modified by
weatheringprocesses
operatingon the surfaceof the flows as a
function of time. Where there is intensetectonic activity, the
underlyingfresh material is brought to the surfaceand highdielectric material is exposed.In this model the occurrenceof
high-dielectricmaterial at high elevationsis related to the fact
that Theia Mons is being constructedon top of the Beta rise
and that tectonicdeformationtendsto producehigh topography, as in the caseof Maxwell Montes.
If this simple model is the case,then what high-dielectric
materials might be candidates for the observed units? The
range of reflectivity in the high-reflectivityunit could be consistentwith averageterrestrial basaltic rock compositionsbut
with relatively high concentrations(10-15 wt %, for example)
of minerals with metallic elements such as Fe, Ti, Mn, or Pb.
Such rock types would be typical of the reflectivity values
mapped in the westernAtalanta Planitia region, for example.
However, Theia Mons, Maxwell Montes, and Ovda and Atla
regionesare characterizedby reflectivityvaluesthat are above
areas south of -30 ø to the east and west of Lada Terra
the range of values typical of terrestrial rocks (e.g., from 0.25
(Plates 4, 6, and 7). The highest reflectivity values in this unit to 0.45, or e from 9 to 27) [Campbell and Ulrichs, 1969], lunar
occur associatedwith Theia Mons, Maxwell Montes, and Atla
basalts [Olhoeft and Strangway, 1975], and basaltic achonRegio. These areas also show very high roughnessvalues.The drites [Campbelland Ulrichs, 1969]. One possibleexplanation
location, elevation, and geologic characteristicsof these re- is the presenceof very high concentrations(> 12-15 wt %) of
gions provide some constraints on the consideration of hy- metallic oxidesof Ti, Fe, and Mn (such as ilmenite, FeTiO3).
pothesesfor the origin of the high-dielectricmaterial. First, we Another possibleexplanationmight be the presenceof ultramake the simplifyingassumptionthat the high-dielectricma- mafic rocks, many of which are enriched in Fe and Ti (e.g.,
terial forms from a singlespecificprocessand is containedin Nyirangonga mafic volcanicsof the East African Rift [Bell and
the rock materials,and we investigatethe implicationsof this Powell, 1969]). In addition, local extremely high conassumptionfor elevation, mode of occurrence,and age. One centrationsof high-dielectricmaterials, such as the recently
clear manifestation of altitude dependence would be the discoveredseafloorhydrothermal activity, could averageout
temperature/pressure-dependent
chemical reactions described over the larger Pioneer Venus footprint to yield very high
by Florensky et al. [1977], Khodakovskyet al., [1979], Barsu- anomalies.
kov et al., [1980a, b], and Nozette and Lewis [1982]. However,
A more complicated and perhaps more realistic model
the occurrenceof high-dielectricmaterialsat severalelevations would allow for variousfactorsto govern the origin and dis(Beta/Maxwell and Atalanta Planitia for example) argues tribution of the high-dielectricmaterial. Such a model might
against this as the sole factor under this assumption.Highallow for the primary emplacementof high dielectricsin lava
dielectricmaterialson Venus [Pettengill et al., 1982; Ford and flows and subsequent weathering to obscure the highPettengill, 1983] have severalmodes of occurrencein a geo- reflectivity signature, the preferential emplacementof flows
logical context. The concentrationat Theia Mons is directly with high dielectricsin certain geologicalenvironmentssuch
associatedwith a relatively young shield volcano [Campbell et as rift zones,as is seenon earth, and exposureof high dielecal., 1984a] while the concentrationat Maxwell Montes ap- trics by tectonic activity. Secondaryoccurrencesof high dipears to be more related to the extensivedeformation associ- electric materials might arise from erosion and concentration
ated with the banded texture [Campbellet al., 1983], although in lag deposits, the production of high-dielectric material
through various simplereactions[e.g., Florenskyet al., 1977;
a volcanicorigin for the structurecannot be ruled out [Masurskyet al., 1980]. Thus the mode of occurrencemay be con- Nozette and Lewis, 1982; Garyin et al., 1984d], and the possisistentwith both primary emplacementand exposureby fault- bility of widespreadrock coatingsof high-dielectricmaterial
ing and erosion associatedwith steep slopes.The very high such as Fe and Mn, analogousto desertvarnish or deep-sea
roughnessvalues associatedwith theseregionsis also consis- nodules. At the present time there are insufficientdata to estent with theseobservations.A third considerationis age and tablish or rule out any of thesefactors.Although we believe
the effectsof weatheringas a function of time. Although Theia that age may be an important factor, we must await adMons is interpreted by virtually all workers to be of volcanic ditional geochemicaland high-resolutionimaging data in
origin [Saundersand Malin, 1977; McGill et al., 1981; Camp- order to assessage and other factors.
bell et al., 1984a], it is not the only volcanic feature or strucCONCLUSIONS
ture on Venus. Preliminary analyses of Venera 15/16 data
showevidencefor widespreadflowsat severalaltitude levelsin
Pioneer Venus data sets for global topography, surface
the northern hemispherein regions estimated to have an roughness,and reflectivityprovide important information on
averageage of one billion years[Barsukovet al., 1984]. If the the geological and textural characteristicsof the surface of
high-dielectricmaterials are associatedwith primary volcanic Venus. We have analyzed each of these data sets and their
6880
HEAD ET AL.: SURFACE CHARACTERISTICS OF VENUS
270 ø
0ø
I
90 ø
I
0ø-• .• e'••
%• •
-3
•
• ; •
!
-.-•:•,•%.•
•j:•••,?
.••¾--..:•.....-:......-•.......--..--•-••....
.•o,o
- oø
••_
I
180 ø
I
":•...•
•
•
--3oø
i
270 ø
0ø
90 ø
180 ø
Plate 1. Referencemap of Venus showinggeographiclocation of featurescited in the text.
o
-
3-0
0
-.
i
60
1•
(KM)
8
Plate 2. Topography of Venus. Four divisionsin topographyare shown in a Mercator projectionof the planet at
1ø x 1ø resolution.The four divisionsare expressedin kilometersrelativeto a planetaryradiusof 6051.0km: (1) < 6051.0
kin, lowlands(purples),(2) 6051.0-6053.0km, rolling plains,(blues);(3) 6053.0-6055.5km, highlands(yellows),and (4)
> 6055.0 km, mountainous regions(reds).
1','-
-
-'
60
1•0
<m
1: 5
ß
Plate 3. Radar roughnessof the surfaccof Venus. Three divisionsin PV radar roughnessarc shown in the same
projccfio• as Plate 2. The three divisions, given in degrees rms slope, arc (l) 1.0ø-2.5ø, smooth (blues), (2) 2.5ø-5.0ø,
transitional(yellows),and (3) $.0ø-10.0ø,rough (reds).
im
Plate 4. Radar reflectivityof the surfaceof Venus.Three divisionsin PV reflectivityare shownin the sameprojection
as Plates 2 and 3. The three divisionsare (1) 0.02-0.1 (blues),predominantlyporous material such as soil, (2) 0.1-0.2
(yellows),predominantlymaterial comparableto terrestrialrock, and (3) 0.2-0.5 (reds),material with a significantpercentage of a high-dielectriccomponent.
6881
4•
6882
HEAD ET AL.' SURFACE CHARACTERISTICS OF VENUS
KEY
E
L 12.0
V
.....
.•. ..&-• •.'• .
A
2.0
I
0.0
o
M -2.5
:
-
.
K
M
30-,
UJ
0
- '•
20
180
RMS
SLOPE
240
Plate 5. Map of elevation versusroughness.Twelve subdivisionsarc definedon the basisof the divisionsof elevation
and roughnessshownin Plates2 and 3 and discussed
in the text. Huc is usedto showdivisionsin elevationand intensity
to show divisionsin roughness.
E
2
K
M
,
6
2
18
2-0
Plate 6. Map of elevationversusreflectivity.Twelve subdivisions
are definedon the basisof the divisionsof elevation
and reflectivityshownin Plates2 and 4 and discussed
in the text. Hue is usedto showdivisionsin elevation,and intensity
to show divisionsin reflectivity.
HEAD ET AL.' SURFACECHARACTERISTICS
OF VENUS
p-
N
w
.
•
w
•P-
6883
6884
HEAD ET AL..' SURFACECHARACTERISTICS
OF VENUS
relation
to eachotherand'havelocally
calibrated
thegeologi-
by Venera 15 and Venera 16 probes:Preliminarydata (in Russian),
cal significanceof some of the distinctive surface types by
examining high-resolutionimages from Venera lander spacecraft and the Arecibo Observatory. These global data set correlations provide a fundamental framework for the understanding of the nature of the surfaceof Venus and will permit
extrapolation of local and regional findings of future geochemical and imaging experimentresultsto a global context.
Our presentconclusionsand interpretationsare as follows:
1. Regional rock and bedrock surfacesare extremelywidespread on Venus, covering the majority of the planet. We
interpret the relatively smooth rock and bedrock surfacesto
be of volcanic origin, most likely representinglava flows, particularly over wide areas of the lowlands and rolling uplands.
Rougher rock and bedrock surfacesare locally related to tec-
Geokhimiya,12, 1811-1820, 1984.
tonic deformation.
2.
Porous and unconsolidated fine materials (soil) cover
less than about 27% of the surface of Venus. The soil surfaces
are generallypatchyin their distribution,are not preferentially
located around major volcaniccomplexes,are to somedegree
localized along the flanking slopesof the highlands,and do
not preferentially occur in the lowlands. We interpret this to
mean that major contributions to soil formation are not
impact-producedregolith or extensivepyroelasticmantlesbut
rather local weathering and small amounts of lateral transport. If the formation of soil is related to age of surfaces,then
soil-rich areas suchas Lada Terra, at southernhigh latitudes,
may be relatively old.
3. A small portion of the surfaceof Venus is characterized
by materials with a significant amount of high-dielectricmaterial near the surface.Although occurring predominantly at
high altitudes, thesesurfacesdo not appear to be solelyrelated
to temperature/pressure-dependent
chemical reactionsbecause
of their distribution at other elevations.We investigateseveral
hypothesesfor the origin of thesematerials and interpret some
occurrences(e.g, Theia Mons in Beta Regio) to representprimary volcanic rocks enriched in such componentsas metals or
metallic oxides,and others (Maxwell Montes in Ishtar Terra)
to representhigh-dielectricmaterial exposedby tectonicdeformation. A major difficulty in interpreting the origin of the
components producing the high dielectric signatures is the
lack of laboratory measurementsfor a wide range of geological materials at thesewavelengthsfor Venus conditions.These
data are absoluteessentialfor a more thorough understanding
of the primary and secondarygeologicalprocessesoperating
to form and modify the surfaceof Venus.
Acknowledgments. We are particularly grateful to Gordon Pettengill and Peter Ford of MIT for providing updated versionsof the
Pioneer Venus data. We wish to acknowledge the help of M. E.
Murphy in manuscript preparation; E. Robinson and J. Tingle for
computer programming support; and R. Grieve, B. Jakosky, R. Simpson, and two anonymous reviewers for manuscript review. This research was carried out under NASA grant NGR-40-002-088 to
J.W.H.
REFERENCES
Barsukov, V. L., I. L. Khodakovsky, V. P. Volkov, and C. P. Florensky, The geochemicalmodel of the troposphereand lithosphere
of Venus based on new data, SpaceRes.,20, 197-203, 1980a.
Barsukov, V. L., V. P. Volkov, and I. K. Khodakovsky, The mineral
compositionof surfacerocks, A preliminary prediction, Proc. Lunar
Planet. Sci., 11th, 765-773, 1980b.
Barsukov, V. L., A. T. Basilevsky,R. O. Kuzmin, A. A. Pronin, V. P.
Kryuchkov, O. V. Nikolaeva, I. M. Chernaya, G. A. Burba, N. N.
Bobina, V. P. Shashkina, M. S. Markov, and A. L. Sukhanov,
Geology of Venusfrom the resultsof analysisof radar imagestaken
Basilevsky,A. T., N. N. Bobina,V. P. Shaskina,Y. G. Shkuratov,Y.
V. Kornienko,A. Y. Usikov,and D. G. Stankovich,Analysisof the
relationshipbetweenaltitude and degreeof surfaceroughness,
Moon Planets,27, 63-89, 1982.
Bell, K., and J. L. Powell, Strontiumisotopestudiesof alkalic rocks:
The potassiumrich lavasof the Berungerand Toro-Ankoleregions,
eastand central equatorial Africa, J. Petrol., 10, 536, 1969.
Campbell,D. B., and B. A. Burns,Earth-basedradar imageryof
Venus, J. Geophys.Res.,85, 8271-8281, 1980.
Campbell,D. B., J. W. Head, J. K. Harmon, and A. A. Hine, Venus:
Identificationof banded terrain in the mountainsof Ishtar Terra,
Science,221, 644-647, 1983.
Campbell, D. B., J. W. Head, J. K. Harmon, and A. A. Hine, Venus:
Volcanismand rift formationin Beta Regio,Science,226, 167-170,
1984a.
Campbell, D. B., J. K. Harmon, A. A. Hine, and J. W. Head, The
surface of Venus: Recent radar observations(abstract), Lunar
Planet. Sci.,XV, 120, 1984b.
Campbell, M. J., and J. Ulrichs, Electrical propertiesof rocks and
their significance
for lunar radar observations,
J. Geophys.
Res.,74,
5867-5881, 1969.
Carmichael,R. S., Handbookof the PhysicalPropertiesof Rocks,vol.
I, 404 pp., CRC Press,Boca Raton, Fla., 1982.
Davis, P. A., and G. G. Schaber, Global units on Venus derived from
statisticalanalysisof PioneerVenusradar data, Lunar Planet. Sci.,
XV, 196-197, 1984.
Ehmann, W. J., and J. W. Head, Aphrodite Terra, Venus:Characteristicsof geologicprovinces(abstract),Lunar Planet.Sci.,XIV, 171172, 1983.
Florensky,C. P., L. B. Ronca, A. T. Basilevsky,G. A. Burba, O. V.
Nikolaeva, A. A. Pronin, A.M. Trakhtman, V. P. Volkov, and V.
V. Zazetsky,The surfaceof Venusas revealedby SovietVenera 9
and 10, Geol.Soc.Am. Bull., 88, 1537-1545, 1977.
Florensky,C. P., V. P. Volkov, and O. V. Nikolaeva,A geochemical
model of the Venus troposphere,Icarus,33, 537-553, 1978.
Florensky, C. P., A. T. Basilevsky,G. A. Burba, O. V. Nikolaeva, A.
A. Pronin, A. S. Selivanov,M. K. Narayeva, A. S. Panfilov, and V.
P. Chemodanov, Panoramas of Venera 9 and 10 landing sites,in
Venus,edited by D. M. Hunten, L. Colin, T. M. Donahue, and V. I.
Moroz, Chap. 8, p. 137, University of Arizona Press,Tucson, 1983.
Ford, P. G., and G. H. Pettengill, Venus: Global surfaceradio emissivity,Science,220, 1379-1381, 1983.
Ford, P. G., and G. H. Pettengill, Venus radar reflectivity--a reanalysis,Bull. Am. Astron.Soc.,16, 696, 1984.
Garvin, J. B., J. W. Head, and L. Wilson, Magma vesiculationand
pyroclasticvolcanismon Venus,Icarus, 52, 365, 1982.
Garvin, J. B., J. W. Head, S. H. Zisk, and G. H. Pettengill, Venus
global radar roughness:A preliminary analysis (abstract), Lunar
Planet. Sci., XIV, 239-240, 1983a.
Garvin, J. B., J. W. Head, A. R. Peterfreund, G. H. Pettengill, and S.
H. Zisk, Venus: Global distribution of radar roughnessand reflectivity, Bull. Am. Astron.Soc.,15, 818, 1983b.
Garvin, J. B., J. W. Head, G. H. Pettengill, and S. H. Zisk, Venus:
Global distribution of radar roughnessand reflectivity and correlations with elevation (abstract), Lunar Planet. Sci., XV, 294-295,
1984a.
Garvin, J. B., J. W. Head, and A. T. Basilevsky,Venera lander site
geologic characteristicsfrom Pioneer-Venus radar measurements
(abstract),Lunar Planet. Sci.,XV, 292-293, 1984b.
Garvin, J. B., J. W. Head, M. T. Zuber, and P. Helfenstein, Venus:
The nature of the surfacefrom Venera panoramas,J. Geophys.Res.,
89, 3381-3399, 1984c.
Garvin, J. B., R. A. Fogel, R. A. F. Grieve, and J. W. Head, A comparison of theoreticalmineral equilibria and Pioneer-Venusradar
data (abstract),Lunar Planet. Sci.,XV, 288-289, 1984d.
Garvin, J. B., J. W. Head, G. H. Pettengill, and S. H. Zisk, Venus
global radar reflectivityand correlationswith elevation,J. Geophys.
Res., 90, 6859-6871, 1985.
Greeley, R., J. Iversen, R. Leach, J. Marshall, B. White, and S. Williams, Windblown sand on Venus: Preliminary reults of laboratory
simulations, Icarus, 57, 112-124, 1984.
Hagfors, T., Backscatteringfrom an undulating surface with applications to radar returns from the moon, J. Geophys.Res., 69, 37793784, 1964.
Hagfors, T., Remote probing of the moon by infrared and microwave
emissionsand by radar, Radio Sci., 5, 219-227, 1970.
HEAD ET AL.: SURFACE CHARAC•ISTICS
Head, J. W., and L. Wilson, Volcanic processeson Venus (abstract),
Lunar Planet. $ci., XIII, 312-313, 1982.
Head, J. W., D. B. Campbell, and S. H. Zisk, Tectonism and volcanism on the southern slopes of Ishtar Terra, Venus (abstract),
Lunar Planet. $ci., XVI, 331-332, 1985.
Khodakovsky, I. L., V. P. Volkov, I. Y. Sidorov, M. V. Borisov, and
M. V. Lomonosov, Venus: Preliminary prediction of the mineral
compositionof surfacerocks,Icarus, 39, 352-360, 1979.
Krotikov, V. D., Several electrical characteristics of the earth's rocks
and a comparison with the surfacelayer of the moon, Izv. Vuzov.
Radiofiz.,5/6, 1057-1061, 1962.
Krotikov, V. D., and V. S. Troitsky, Radio emissionand the nature of
the moon, Usp. Fiz. Nauk., 81, 589-639, 1963.
Masursky, H., E. Eliason, P. G. Ford, G. E. McGill, G. H. Pettengill,
G. G. Schaber, and G. Schubert, Pioneer-Venus radar results: Geology from images and altimetry, J. Geophys.Res., 85, 8232-8260,
1980.
Masursky, H., M. E. Strobell, and K. E. Beer, Planetary nomenclature, NASA Tech. Mem., TM-86246, 353-355, 1984.
McGill, G. E., S. J. Steenstrup,C. Barton, and P. G. Ford, continental
rifting and the origin of Beta Regio, Venus, Geophys.Res. Lett, 8,
737-740, 1981.
McGill, G. E., J. L. Warner, M. C. Malin, R. E Arvidson, E. Eliason,
S. Nozette, and R. D. Reasenberg,Topography, surfaceproperties,
and tectonicevolution, in Venus,edited by D. M. Hunten, L. Colin,
T. M. Donahue, and V. I. Moroz, Chap. 6, p. 69, University of
Arizona Press, Tucson, 1983.
Nozette, S., and J. S. Lewis, Venus: Chemical weathering of igneous
rocks and buffering of atmosphericcomposition,Science,216, 181183, 1982.
Olhoeft, G. R., and D. W. Strangway,Dielectric propertiesof the first
100 meters of the moon, Earth Planet. $ci. Lett., 24, 394-404, 1975.
Ostro, S. J., Planetary radar astronomy, Rev. Geophys.,21, 186-196,
1983.
Parkhomenko, E. I., Electrical Propertiesof Rocks, 314 pp., Plenum,
New York, 1967.
OF VENUS
Pettengill, G. H., E. Eliason, P. G. Ford, G. B. Loriot, H. Masursky,
and G. E. McGill, Pioneer-Venusradar results:Altimetry and surface properties,J. Geophys.Res.,85, 8261-8270, 1980b.
Pettengill, G. H., P. G. Ford, and S. Nozette, Venus: Global surface
radar reflectivity, Science,217, 640-642, 1982.
Saunders, R. S., and M. C. Malin, Geologic interpretation of new
observationsof the surfaceof Venus, Geophys.Res. Lett., 4, 542550, 1977.
Schaber, G. G., R. C. Kozak, P. Davis, and E. Eliason, Venus Pio-
neer: Ratios and composite maps of altimetry, RMS slope, and
Fresnel reflection coefficient (abstract), Lunar Planet. Sci., XIII,
683-684, 1982.
Sharpton, V. L., and J. W. Head, Regional slope characteristicsof
global topography: A comparison of Venus and earth (abstract),
Lunar Planet. Sci., XV, 760-761, 1984.
Simpson,R. A., G. L. Tyler, and G. G. Schaber,Viking bistatic radar
experiment:Summary of resultsin near-equatorialregions,J. Geophys.Res.,89, 10,385-10,404, 1984.
Strobell, M. E., and H. Masursky, Venus nomenclature and mythology, Pioneer Venus, edited by R. Firereel, L. Colin, and E. Burgess,
NASA Spec.Publ., SP-461, 201-205, 1983.
Tyler, G. L., and H. J. Howard, Dual-frequencybistatic-radar investigations of the Moon with Apollos 14 and 15, J. Geophys.Res., 76,
4852-4874, 1973.
U.S. Geological Survey,Altimetric and shadedrelief map of Venus, in
Atlas of Venus, Map 1-1324, scale 1: 50,000,000 Reston, Va., 1981.
U.S. Geological Survey, Topographic and shaded relief maps of
Venus, Map 1-1562, scale 1:50,000,000, Reston, Va., 1984.
White, B. R., Venusian saltation, Icarus, 46, 226-232, 1981.
J. B. Garvin, GeophysicsBranch, NASA Goddard Space Flight
Center, Code 622, Greenbelt, MD 20771.
J. W. Head III and A. R. Peterfreund,Department of Geological
Sciences,Brown University, Providence,RI 02912.
S. H. Zisk, MIT/NEROC Haystack Observatory, Westford, MA
01886.
Pettengill, G. H., Physical propertiesof the planets and satellitesfrom
radar observations, Annu. Rev. Astron. Astrophys., 16, 265-292,
1978.
Pettengill, G. H., D. F. Horwood, and C. H. Keller, Pioneer-Venus
orbiter radar mapper: Design and operation, IEEE Trans. Geosci.
Remote $ens., GE-18, 28-32, 1980a.
6885
(ReceivedDecember 7, 1983;
(revisedFebruary 7, 1985;
acceptedMarch 5, 1985.)