GEOLOGY AND TOPOGRAPHY OF RA PATERA, IO, IN THE VOYAGER ERA:
		PRELUDE TO ERUPTION

		Paul M. Schenk1, Alfred McEwen2, Trevor Davenport3, 
		A. G. Davies4, Kevin Jones5, and Brian Fessler1 

		1Lunar and Planetary Institute, Houston, TX
		2University of Arizona, Tucson, AZ
		3South Dakota School of Mines, Rapid City, SD
		4Jet Propulsion Laboratory, Pasadena, CA
		5Brown University, Providence, RI


Correspondence author:
Paul Schenk
Lunar and Planetary Institute, Houston, TX 77058
281 486-2157
schenk@lpi.usra.edu

Submitted to Geophysical Research Letters, April 4, 1997

ABSTRACT

 Ra Patera is a major active volcanic center on Io and a controversial site of liquid sulfur eruptions.  Voyager stereo 
images are used to map the geology and topography of Ra Patera as of 1979 in order to characterize factors that 
influenced eruptions observed in 1994-95 (Spencer et al., 1997).  The summit of Ra Patera reaches a height of only 
~1 km.  Voyager-era flows, up to 250 km in length, formed on slopes averaging 0.1 to 0.3¡, comparable to those on 
the lunar mare.  The only significant positive relief near Ra Patera is a 600-km-long mesa and mountain unit that 
reaches a maximum height of ~8 km near Carancho Patera.  This unit also forms a 60 by 90 km wide plateau ~0.5 
to 1 km high extending to within 50 km of Ra Patera.  The broad dark flow observed by Galileo (Belton et al., 
1996;  McEwen et al., 1997) extending southeast of the Ra Patera caldera apparently flowed around the southwestern 
edge of the plateau on slopes of ~0.2 to 0.3¡.  These parameters may place important constraints on the rheology and 
perhaps the composition of volcanic flows at Ra Patera.

Introduction
	Most volcanic flows are influenced by local and regional topography and the numerous 
eruptions on Io should be no exception.  Ra Patera (-8¡, 325¡) is the largest known of IoÕs radiating 
shield-like lava flow fields (total area ~250,000 km2;  Fig. 1).  It is a controversial site of sulfur or 
sulfur-rich lava flows (Pieri et al., 1984;  McEwen et al., 1989; Greeley et al., 1990;  Moses and 
Nash, 1991) and has been the scene of some of the most dramatic surface changes observed over 
the 17 years since the 1979 Voyager encounters.  HST observed a major brightening at Ra between 
March 1994 and July 1995 (Spencer et al., 1997).  Galileo images showed an active plume, 
extensive diffuse bright deposits, and a large dark deposit interpreted to be a massive lava flow or 
flow field extending southeast from the central vent area (Belton et al., 1996).  Our goal is to 
characterize pre-eruption factors that influenced the shape and location of new lava flows and 
deposits.  The style and sequence of past volcanic events at Ra Patera may also provide insights 
into more recent and future volcanic activity.  We have used recently processed 1979 Voyager 
stereo images to map the pre-1994 geology and topography of Ra Patera.     

Voyager 1 stereo coverage
	Voyager obtained redundant image coverage of the Ra Patera region at four distinct viewing 
(or phase) angles.  Voyager FDS images 16390.06/10 and 16392.57/59 (Fig. 1) provide the most 
useful stereo coverage for both geologic (Fig. 2) and topographic (Fig. 3) mapping, with an 
effective ground resolution of 1.6 km/pixel and a stereo phase angle of 50¡.  Topography was 
mapped using automated stereogrammetry correlation software developed at LPI from related 
PICS/ISIS (USGS, Flagstaff) software.  The software uses a scene recognition algorithm to locate 
features in each of the stereo images at ~1/5th pixel accuracy, with a nominal vertical resolution of 
210 m for the images used here.  The observed relative displacement of each feature is a measure 
of the parallax, from which height is calculated.  These relative heights are mapped across the 
scene as a digital elevation model (DEM) (Fig. 3).    
	Several problems effect stereo topography mapping on Io in particular.  Some volcanic 
materials on Io have very different photometric curves or visible colors.  The large stereo phase 
angle difference (50¡) and use of images taken through different filters (clear and green in this case) 
result in contrast reversals within many of the small dark spots on Io.  Diffuse bright deposits also 
appear to become more or less prominent depending on viewing geometry and image filter.  These 
effects will confuse the scene matching algorithm, but are easily detected visually in the stereo 
image model and areas affected have been removed manually from our DEM.  An additional  
problem is the extensive occurrence of smooth plains on Io (Schaber, 1982).  The lack of discrete 
features within these plains at Voyager resolution results in large areas of noisy or missing data in 
the DEM, which must also be removed (Fig. 3).
		In the absence of a global topographic model at sufficient resolution, we assume that the 
regional gradient across Ra Patera, which is ~450 km across, is approximately zero.  Control nets 
for each image pair were updated, and our DEMÕs were adjusted where necessary to remove any 
regional gradients that might be due to inadequate image registration.  Galileo may ultimately 
provide a useful global topographic datum that can be used to more precisely control our higher 
resolution topographic models.    
	The sampling resolution of the DEM can be varied to emphasize topography at different 
wavelengths.  For Ra Patera a topographic resolution cell of 33 km (21 pixels) was chosen to 
emphasize large scale relief and overall structure and to maximize signal-to-noise (due to the 
presence of diffuse and featureless deposits within Ra).  A resolution cell of 8 km (5 pixels) was 
chosen for the plateau and mountain unit associated with Carancho Patera (discussed below) to 
emphasize smaller-scale detail in that area.  Finally, the DEM was cross-correlated visually with the 
stereo image model, which reveals several unexpected features.

 Geology of Ra Patera
	The radiating dark flows observed by Voyager at Ra Patera were formed on a mottled plains 
unit that is smooth and relatively bright with numerous dark spots (Figs. 1, 2).  This unit may be 
comprised of multiple overlapping but unresolved volcanic deposits (Greeley et al., 1988), and is 
the oldest unit identified.  Surrounding the Ra Patera volcanic field is an extensive smooth plains 
unit (Figs. 1, 2) that is darker than the mottled plains and generally featureless at Voyager 
resolution (Greeley et al., 1988).  An outward facing scarp is visible in a few locations.  These 
plains may be massive sheet flows or innumerable smaller overlapping flows that were not deep 
enough to bury the Ra Patera edifice.  Numerous dark-floored caldera occur thorughout these 
plains.  The dark Ra Patera longitudinal flows emanating from the dark 30-km-long central vent 
may be either sulfur-rich (Pieri et al., 1984) or silicic in composition.  These flows appear to 
postdate both the smooth and mottled plains units.  They are between ~1 km (the limit of 
resolution) and 4 km wide, and are ~50 to 250 km long.  In one area, flows appear to coalesce or 
spread out into a broader flow field ~15 km across.  At numerous sites, small lobes appear to 
branch laterally from the main flow (Greeley et al., 1988).
	Smooth mesa material is topographically thick and has sharp outward facing scarps.  These 
occur adjacent to Huo Shen Patera and at a small feature northwest of Ra Patera (Fig. 2).  These 
may be viscous lava extrusions or erosional remnants.  Mountain material consists of elevated 
plateau and ridge material and occurs at a 100 km long unnamed mountain due northwest of Ra 
Patera and in 600-km-long arcuate structure near Carancho Patera that extends to within 50 km of 
the Ra summit (Fig. 2).  The small unnamed mountain is striated and probably tectonic in origin.  
The arcuate structure near Carancho Patera is more extensive and massive than previously thought.  
This feature has an average width of almost 100 km and reaches to within 50 km of the center of 
Ra Patera, where it forms a scarp-bounded plateau 60 by 90 km across.  The relative ages of the 
smooth mesa materials and the mountain materials are uncertain but appear to postdate both smooth 
and mottled plains.  The youngest pre-1994 unit is diffuse mantling material (Fig. 2).  The contact 
of this unit is gradational over 20 km distance and is distinctly reddish (check) relative to older 
units.  This unit crosses but does not obscure preexisting contacts, indicating it is not 
topographically thick but is a thin covering, and is probably a late-stage plume-like deposit.

Topography of Ra Patera
	Very little relief is observed across Ra Patera, despite the 210 m vertical resolution of our 
DEM.  Individual flows are not resolved in the DEM, giving us an upper limit on their thickness of 
~200 m.  The summit of Ra Patera rises only ~1.0±0.2 km above the surrounding dark smooth 
plains.  Slopes across the mottled plains, on which the 250-km-long dark longitudinal flows 
formed, average 0.1 to 0.2¡.  Slightly higher slopes of 0.2 to 0.3¡ are observed to the southeast of 
the central caldera.  The only substantial relief within the Ra Patera flow field is associated with the 
600-km-long mesa and mountain structure extending between Ra Patera and Carancho Patera 
(Figs. 1, 2, 3).  Stereo images reveal that these units are  0.5 to 2.0 km high (Figs. 1, 2). The 
plateau 50 km east of Ra Patera is a potential topographic impediment to any eastward lateral flow 
of lava.
	The highest relief in the region occurs within the unnamed oval promontory ~50x100 km 
across just northwest of Carancho Patera (Fig. 1, 2) which is ~8 km high.  (Only Haemus Mons (9 
km) and Euboea Montes (10.5 km) are higher [Schenk and Bulmer, 1997].  The smaller unnamed 
mountain northwest of Ra is 4-5 km high.)  Carancho Patera itself is only 1 to 1.5 km above the 
surrounding plains and may not be the direct source of the deposit.  A 10-km-wide oval pit near the 
summit of the mountain (FDS 16390.56) may be the source vent.  South of Carancho Patera, the 
mountain unit is 2 to 3 km high and is banded or striated parallel to the margin of the deposit, 
suggesting a volcanic or constructional in origin (Greeley et al., 1988).  Several transverse 
structures ~5 to 6 km high cross this striated deposit south of Carancho Patera (Figs. 1, 3).  These 
may be vent sources.   

Discussion 
	The slopes observed at Ra Patera (<0.3¡) are very low compared to most terrestrial shield 
volcanos, such as Kilauea, Mauna Loa, and Olympus Mons on Mars (Moore et al., 1978), and 
many Venus volcanos (Schaber, 1991).  Alba Patera, Mars (Mouginis-Mark et al., 1988), and a 
few venusian shields (Schaber, 1991) have slopes as low as 0.2¡.  The compositions of these 
volcanos are usually assumed to be basaltic.  The slopes on Ra Patera are also similar to those on 
basaltic plains such as Mare Imbrium (Moore and Schaber, 1975), portions of the Snake River 
plains (Greeley and King, 1977), and a few large flow fields on Venus (Roberts et al., 1992).  
Low slopes do not disallow the eruption of either sulfur or silicate lavas at Ra Patera but may 
constrain the nature of eruption.  The basaltic Imbrium flows (up to 400 km long) are  inferred to 
have formed at very low viscosity (due to low silica and/or high metal content; Carr, 1973) or very 
high eruption rates (Schaber, 1973).  By inference, the Voyager-era flows on Ra Patera, if silicate, 
might also be characterized as of low viscosity, high eruption rate or both.  High eruption rates 
might lead to turbulent flow which could allow for the longer runout lengths over shallow slopes 
seen on the Ra patera flows.  
	Davies (1996) has applied the Bingham-type model of Hulme (1974) to ionian eruption 
conditions.  Using this model and the new slope measurements, we can estimate silicate flow 
dimensions and eruption parameters for the Ra Patera flows observed by Voyager.  Flow channel 
width is proportional to the mass eruption rate and inversely proportional to the underlying slope 
(see Wilson and Head, 1983).  For a basaltic magma (see Davies, 1996, for flow model 
parameters), with a viscosity of 1000 Pa s, and a slope of 0.3¡, the expected channel width for a 2-
km-wide flow with levees on either side is 1200 m.  A mass eruption rate inferred from channel 
width (Davies, 1996) is about 60 m3 s-1.  Such a flow would have a central depth of ~6.5 m and 
advance very slowly (~10 m2 s-1).  For a 5-km-wide flow, the channel width is 4200 m, flow 
depth is 9.4 m, and mass eruption rate is 2000 m3 s-1.  This flow would take ~40 days to reach 
250 km (280 days for the smaller width flow).  A change of width from 2 to 5 km could be 
achieved by a simple change in slope of as little as 0.1¡.  
	The mass eruption rates implied for the Ra flows are considerably less than those associated 
with the large 4.8 micron thermal outbursts associated with Io:  the Loki outburst of January 1990 
has been modeled with mass eruption rates of 105 m3 s-1 (Davies, 1996), in the range of the lunar 
mare basalt emplacement rates (e.g., Schaber, 1973).  On Earth, the largest mass eruption rate 
observed is for the Laki, Iceland eruption of 1783.  Basalt from a 25 km long fissure erupted for 7 
months, for the first two months at a rate of 0.1 km3 per day: close to the 2000 m3 s-1 (0.17 km3 
per day) calculated for the Ra Patera flows.  Indeed, there are striking geomorphological 
similarities in the shapes and scales of the Laki and Ra Patera flows.
	The new lava flows observed by Galileo (Belton et al., 1996) appear to be much broader and 
shorter in length than the longitudinal Voyager-era flows (Fig. 4).  This new deposit may be one 
massive flow or a consolidated flow field comprised of numerous small flows.  There appears to 
have been a change in volcanic style from long narrow longitudinal flows to shorter and broader 
flows at Ra Patera.  Although measured slopes in this region are a factor of 2 higher than 
elsewhere on Ra Patera, they are still <0.3¡.  Ra Patera may have been originally much higher (i.e., 
>3 km) and steeper (i.e., >0.5¡) but subsided after the formation of the longitudinal flows due to 
volcanic deflation or lithospheric mass loading.  No evidence of concentric fracture patterns or 
topographic flexure or swells is observed near Ra Patera.  Also, the lithosphere may be ~30 km 
thick (Nash et al., 1986), consistent with the support of 8 km of relief near Carancho, and we 
conclude that significant deflation of Ra Patera was unlikely.  Alternatively, there may be a 
depression in the area where the new flows formed, causing the flow(s) to spread laterally and 
pond.  A shallow depression in this area is suggested but not confirmed by the Voyager DEM.  

Conclusions
	Longitudinal lava flows on Ra Patera observed by Voyager were followed by at least one 
episode of plume-like eruption and deposition which mantled but did not obscure older units.  A 
prominent 1-3 km high plateau ~50 km due east of summit and extending 600 km due northeast 
and then northwest are the only significant positive relief features near Ra Patera.  Indeed, a 
preliminary comparison of the location of the new HST-Galileo dark flow at Ra Patera (Spencer et 
al., 1997;  Belton et al., 1996) indicates that this deposit flowed around the southeast edge of this 
plateau (Fig. 4) in a region where slopes appear to be slightly steeper, although still <0.3¡, than 
elsewhere on Ra Patera.  These low slopes are similar to those observed on the lunar mare and on a 
few unusually flat shield volcanos on Mars and Venus.  Modeling suggests that silicates can flow 
over such long distances and shallow slopes if eruption rates are relatively high and the lavas have 
relatively low viscosity.  The rheological properties and flow behavior of sulfur may be complex 
(Fink et al., 1983;  Greeley et al., 1990).  Natural leveed sulfur flows up to 1 km long are 
observed (e.g., Watanabe, 1940) and formation of the long Ra flows as sulfur may simply require 
high eruption rates or durations.  Rapid formation of crusts may reduce heat loss to the point where 
flow lengths can be increased dramatically (Greeley et al., 1990) but whether sulfur can flow in 
narrow channeled flows for 250 km over slopes as low as 0.1¡ has not yet been demonstrated.


ACKNOWLEDGEMENTS
We thank Kay Edwards, USGS Flagstaff, for valuable programming assistance.  LPI Contribution 
no. xxx.

REFERENCES

Belton, M., et al., GalileoÕs first images of Jupiter and the Galilean satellites, Science, 274, 377-385, 1996.
Davies, A., IoÕs volcanism: Thermophysical models of silicate lava compared with observations of thermal 
emission, Icarus, 124, 45-61, 1997.
Fink, J., S. Park, and R. Greeley, Cooling and deformation of sulfur flows, Icarus, 56, 38-50, 1983.
Greeley, R., P. Spudis, and J.Guest, Geologic map of the Ra Patera Area,  USGS Misc. Invest. Series Map I-1949, 
1988.
Greeley, R., S. Lee, D. Crown, and N. Lancaster, Observations of industrial sulfur flows: Implications for Io, 
Icarus, 84, 374-402, 1990.
Hulme, G., The interpretation of lava flow morphology, Geophys. J. R. Astron. Soc., 39, 361-383, 1974.
McEwen, A., J. Lunine, and M. Carr, Dynamic geophysics of Io, in Time-variable phenomena in the Jovian 
system, NASA SP-494, p. 11-46, 1989.
McEwen, A., et al., 17 years of surface changes on Io: Galileo SSI results, Lunar Planet. Sci. XXVIII, 907-908, 
1997.
Moore, H., and G. Schaber, An estimate of the yield strength of the Imbrium flows, Proc. Lunar Planet. Sci. Conf., 
6th, 101-118, 1975.
Moore, H., D. Arthur, and G. Schaber, Yield strengths of flows on the Earth, Mars, and Moon, Proc. Lunar Planet. 
Sci. Conf., 9th, 3351-3378, 1978.
Moses, J, and D. Nash, Phase transformations and the spectral reflectance of sold sulfur: Can metastable sulfur 
allotropes exist on Io? Icarus, 89, 277-304, 1991.
Mouginis-Mark, P., L. Wilson, and J. Zimbelman, Polygenic eruptions on Alba Patera, Mars, Bull. Volcanol., 50, 
361-379, 1988.
Nash, D., M. Carr, J. Gradie, D. Hunten, and C. Yoder, Io, in Satellites, Univ. Arizona Press, Tucson, pp. 629-
688, 1986.
Pieri, D., S. Baloga, R. Nelson, and C. Sagan, Sulfur flows of Ra Patera, Io, Icarus, 60, 685-700, 1984.
Roberts, K., J. Guest, J. Head, and M. Lancaster, Mylitta Fluctus, Venus: Rift-related centralized volcanism and the 
emplacement of large-volume flow units, J. Geophys. Res., 97, 15991-16015, 1992.
Schaber, G., Lava flows in Mare Imbrium: Geological evidence from Apollo orbital photography, Proc. Lunar 
Planet. Sci. Conf. 4, 73-92, 1973.
Schaber, G., The geology of Io, in Satellites of Jupiter, (D. Morrisom, ed.), pp. 556-597, 1982.
Schaber, G., Volcanism on Venus as inferred from the morphometry of large shields, Proc. Lunar Planet. Sci. 
Conf., 21, 3-11, 1991.
Schenk, P., and M. Bulmer, Slope failures at Euboea Montes, Io, Lunar Planet. Sci. XXVIII, 1245-1246, 1997.
Spencer, J., A. McEwen, M. McGrath, P. Sartoretti, D. Nash, K. Noll, and D. Gilmore, Volcanic resurfacing of Io: 
Post-repair HST images, in press, Icarus, 1997.
Watanabe, T, Eruptions of molten sulfur from the Siretoko-Iosan volcano, Hokkaido, Japan, Jap. J. Geol. Geogr., 
17, 289-310, 1940.
Wilson, L., and J. Head, A comparison of volcanic eruption processes on Earth, Moon, Mars, Io and Venus, Nature, 
302, 663-669, 1983.

FIGURES

Figure 1.  Stereo image pair of Ra Patera.  Several dark longitudinal lava flows extend up to ~250 
km from the summit of Ra Patera (oval dark area at lower left center).  Ra Patera itself has very 
little relief.  The only exceptional relief occurs along a complex mesa and mountain unit east and 
northeast of Ra Patera.  The convergence angle for this stereo pair is 50¡, base-to-height ratio ~1.5, 
and vertical exaggeration ~7.5.  North is to the right, and image resolution is 1.35 km/pixel.  

Figure 2.  Geologic sketch map of Ra Patera region.

Figure 3.  Color-coded topographic map of Ra Patera (red=high, blue=low).  The prominent 
plateau unit just below Ra Patera is inadequately represented due to the large sampling area (33x33 
km) used here to map topography.  

Figure 4.  Comparison of the new Ra Patera flow location from Galileo (Belton et al., 1996) with 
geologic map (Fig. 2) based on Voyager images.  Compare the location of the dark flow in image 
center and the location of mesa and mountain units (yellow in map) near Ra Patera.  Scene width is 
~1000 km.