Io Volcanism Seen by New Horizons: A Major Eruption of the Tvashtar Volcano

Abstract

Jupiter's moon Io is known to host active volcanoes. In February and March 2007, the New Horizons spacecraft obtained a global snapshot of Io's volcanism. A 350-kilometer-high volcanic plume was seen to emanate from the Tvashtar volcano (62 degrees N, 122 degrees W), and its motion was observed. The plume's morphology and dynamics support nonballistic models of large Io plumes and also suggest that most visible plume particles condensed within the plume rather than being ejected from the source. In images taken in Jupiter eclipse, nonthermal visible-wavelength emission was seen from individual volcanoes near Io's sub-Jupiter and anti-Jupiter points. Near-infrared emission from the brightest volcanoes indicates minimum magma temperatures in the 1150- to 1335-kelvin range, consistent with basaltic composition.

Full text

NH Long Range Reconnaissance Imager (LORRI)
images (17). This same feature appears in the HST/
SBC image in Fig. 2B, obtained when East Girru
was shifted just behind the limb. The auroral feature
near East Girru in both LORRI and HST/SBC im-
ages is ~15° northward of Jupiter’s field line tangent
point at the limb, which suggests that ionospheric
currents are diverted northward from this nominal
position toward a region of higher gas density near
the plume. Similar deviations of the anti-jo vian FUV
emissions from nominal tangent points observed
with STIS are likely caused by the prevalence and
distribution of plumes there (21). Volcanic plume
aurorae were not identified in previous lower-quality
STIS FUV images, which caused an apparent dis-
crepancy with visible images of plume aurorae. The
East Girru plume FUV auroral feature in Fig. 2 re-
solves this discrepancy and reveals the influence of
plumes on Io’s electrodynamic interaction. The
upstream-side emission feature is more apparent
when limb brightened at viewing geometries like
those reported in Fig. 2. This feature was predicted
by aurora image s imulations (22) and is diagnostic of
the diverg ence of the plasma flow upstream of Io, a
primary trait of Io’s interaction with th e plas ma torus.
References and Notes
1. F. Bagenal, J. Atmos. Sol. Terres. Phys. 69, 387 (2007).
2. E.Lellouch,M.A.McGrath,K.L.Jessup,inIo After Galileo
(Springer-Praxis, UK, 2006), pp. 231–264.
3. J. Pearl et al., Nature 280, 755 (1979).
4. W. M. Sinton, C. Kaminski, Icarus 75, 207 (1988).
5. J. R. Spencer et al., Science 288, 1198 (2000).
6. A. F. Cook et al., Science 211, 1419 (1981).
7. J. T. Clarke, J. Ajello, J. Luhmann, N. M. Schneider,
I. Kanik, J. Geophys. Res. 99, 8387 (1994).
8. P. E. Geissler et al., Science 285, 870 (1999).
9. A. H. Bouchez, M. E. Brown, N. M. Schneider, Icarus 148,
316 (2000).
10. P. E. Geissler et al., J. Geophys. Res. 106, 26137 (2001).
11. P. E. Geissler et al., Icarus 172, 127 (2004).
12. I. DePater et al., Icarus 156, 296 (2002).
13. K. D. Retherford, thesis, The Johns Hopkins University (2002).
14. J. Saur, D. F. Strobel, Icarus 171, 411 (2004).
15. S. A. Stern et al., in Astrobiology and Planetary Missions,
R. B. Hoover, G. V. Levin, A. Y. Rozanov, G. R. Gladstone,
Eds. (Proc. SPIE, Volume 5906, 2005), pp. 358–367.
16. K. D. Retherford et al., paper presented at the
Magnetospheres of the Outer Planets Meeting, San
Antonio, TX, 25 June, 2007.
17. J. R. Spencer et al., Science 318, 240 (2007).
18. The angular size of Io varies with spacecraft distance but
is smaller than the Alice slit width for these data. The
spectral resolution varies between 0.3 nm and ~0.9 nm
for emissions known to be located near the satellite disk
(22); see, e.g., Fig. 2.
19. H. C. Ford et al., in Future EUV/UV and Visible Space
Astrophysics Missions and Instrumentation, J. C. Blades,
O. H. W. Siegmund, Eds. (Proc. SPIE, Volume 4854,
2003), pp. 81–94.
20. F. L. Roesler et al., Science 283, 353 (1999).
21. K. D. Retherford et al., J. Geophys. Res. 105, 27157 (2000).
22. J. Saur, F. M. Neubauer, D. F. Strobel, M. E. Summers,
Geophys. Res. Lett. 27, 2893 (2000).
23. K. D. Retherford, H. W. Moos, D. F. Strobel, J. Geophys.
Res. 108, 1333 (2003).
24. L. M. Feaga, thesis, The Johns Hopkins University (2005).
25. K. L. Jessup et al., Icarus 169, 197 (2004).
26. J. R. Spencer et al., Icarus 176, 283 (2005).
27. P. D. Feldman et al., Geophys. Res. Lett. 27, 1787 (2000).
28. D. F. Strobel, B. C. Wolven, Astrophys. Space Sci. 277,
271 (2001).
29. F. Neubauer, J. Geophys. Res. 103, 19843 (1998).
30. F. M. Neubauer, J. Geophys. Res. 104, 3863 (1999).
31. We thank the New Horizons mission and science teams.
New Horizons is funded by NASA. Support for this work was
also provided by NASA though grant number GO-10871
from the Space Telescope Science Institute (STScI), which is
operated by the Association of Universities for Research in
Astronomy, Inc., under NASA contract NAS5-26555.
Supporting Online Material
www.sciencemag.org/cgi/content/full/318/5848/237/DC1
Figs. S1 and S2
Tables S1 and S2
10 July 2007; accepted 19 September 2007
10.1126/science.1147594
REPORT
Io Volcanism Seen by New Horizons:
A Major Eruption of the
Tvashtar Volcano
J. R. Spencer,
1
* S. A. Stern,
2
A. F. Cheng,
2
H. A. Weaver,
3
D. C. Reuter,
4
K. Retherford,
5
A. Lunsford,
4
J. M. Moore,
6
O. Abramov,
1
R. M. C. Lopes,
7
J. E. Perry,
8
L. Kamp,
7
M. Showalter,
9
K. L. Jessup,
1
F. Marchis,
9
P. M. Schenk,
10
C. Dumas
11
Jupiter’s moon Io is known to host active volcanoes. In February and March 2007, the New Horizons
spacecraft obtained a global snapshot of Io’s volcanism. A 350-kilometer-high volcanic plume was seen to
emanate from the Tvashtar volcano (62°N, 122°W), and its motion was observed. The plume’s
morphology and dynamics support nonballistic models of large Io plumes and also suggest that most
visible plume particles condensed within the plume rather than being ejected from the source. In images
taken in Jupiter eclipse, nonthermal visible-wavelength emission was seen from individual volcanoes near
Io’s sub-Jupiter and anti-Jupiter points. Near-infrared emission from the brightest volcanoes indicates
minimum magma temperatures in the 1150- to 1335-kelvin range, consistent with basaltic composition.
T
he New Horizons (NH) Jupiter flyby pro-
vided the first close-up observations of the
tidally driven volcanism of Jupiter’s moon Io
since the last Galileo orbiter observations of Io in late
2001 (1). The closest approach to Io occurred at
21:57 UT on 28 February 2007 at a range of 2.24
million km. Sunlit observations were made at solar
phase angles from 5° to 159°, and four eclipses of Io
by Jupiter were also observed. NH obtained 190 Io
images with its 4.96 mrad per pixel panchromatic
(400 to 900 nm) Long-Range Reconnaissance Im-
ager (LORRI) and 17 color nighttime and eclipse
images with the 20 mrad per pixel Multicolor V isible
Imaging Camera (MVIC), although MVIC coverage
of Io’s day side was not possible because of detector
saturation. NH also obtained seven 1.25- to 2.5-mm
near-infrared image cubes at 62 mrad per pixel with
the Linear Etalon Infrared Spectral Array instrument
(LEISA) and numerous disk-integrated ultraviolet
observations with the Alice instrument, discussed
separately (2).
Eleven volcanic plumes were identified in the
NH images (Fig. 1A and table S1). In addition to the
single very large “Pele-type” plume at Tvashtar,
which is described separately , NH observed 10 SO
2
-
rich “Prometheus-type” plumes (3–5). These smaller
plumes averaged 80 km high and varied greatly in
brightness. Plumes seen for the first time by NH
include those at Zal and Kurdalagon and a large new
plume, 150 km high, at north Lerna Regio, which
has produced a large albedo change. Three of these
plumes, north Lerna and north and south Masubi,
are associated with recent large lava flows, sup-
porting the idea that Prometheus-type plumes
result from mobilization of surface volatiles by
active lava flows. All active plumes that were on
Table 2. Alice-measured emission line brightness averages and SDs in sunlight and eclipse.
Emission line Type
Disk-average brightness (rayleighs)
IEclipse01 IEclipse04 IEclipse05
OI 130.4 nm Sunlight 706 ± 134 445 ± 27 347 ± 142
Eclipse 597 ± 43 394 ± 14 254 ± 71
Ratio S/E 1.18 ± 0.24 1.12 ± 0.08 1.37 ± 0.68
OI 135.6 nm Sunlight 882 ± 177 577 ± 35 480 ± 188
Eclipse 797 ± 57 536 ± 18 381 ± 94
Ratio S/E 1.11 ± 0.24 1.08 ± 0.08 1.26 ± 0.58
SI 147.9 nm Sunlight 1167 ± 271 986 ± 54 596 ± 288
Eclipse 1205 ± 87 874 ± 28 429 ± 144
Ratio S/E 0.97 ± 0.24 1.13 ± 0.07 1.39 ± 0.82
12 OCTOBER 2007 VOL 318 SCIENCE www.sciencemag.org240
New Horizons at Jupiter
on January 26, 2009 www.sciencemag.orgDownloaded from

the disk or near the limb were also visible in
Jupiter eclipse images because of excitation of
plume gases by the jovian magnetosphere (Fig.
2B), as also seen by Galileo (6).
LORRI imaged almost all of Io at relatively low
phase angles with resolutions between 14 and 22 km
per pixel, providing a surface albedo map suitable for
comparison with previous maps (7) (Fig. 1A). There
are at least 19 locations where surface changes have
occurred since Galileo’s last global images, taken
between 1999 and 2001 (Fig. 1A). The number of
surface changes detected is only one-fourth of those
detected during the 5-year Galileo mission (8), per-
haps because of NH’s lower spatial resolution, the
lack of a color data set with comparable resolution,
and the possibility of surface changes that have faded
since their formation.
The large plume at Tvashtar has renewed the
large ring-shaped plume deposit seen at Tvashtar in
2000 (1), which had been obscured by other plume
deposits by mid-2001. A two-lobed plume deposit
surrounds a new , 240-km-long dark feature, prob-
ably a lava flow, at Masubi (Fig. 1, B to D) created
by the two plumes observed by NH: “North Masubi”
near the vent and “South Masubi” at the distal flow
front. This flow is the longest new lava flow to be
erupted on Io since the 1979 Voyager images. The
North Lerna volcanic plume has produced a 700-km-
wide concentric deposit (Fig. 1, E and F) surround-
ing a fresh, 130-km-long apparent dark lava flow .
Other late-Galileo-era plume deposits, notably around
Dazhbog and Thor (8), have faded to invisibility.
LEISA observed 1.25- to 2.5-mm volcanic ther-
mal emission from Io’s night side or in Jupiter eclipse
at almost all longitudes at a spatial resolution of 140
to 170 km per pixel, producing a uniform global
snapshot of Io’s high-temperature volcanic thermal
emission (Fig. 1A). Thermal emission from several
volcanoes was also seen in 0.4- to 1.0-mmLORRI
images in Jupiter eclipse or on the night side (Fig.
2A). At least 36 hot spots were detected. All cor-
respond to pre viously known active volc anic centers
1
Southwest Research Institute, 1050 Walnut Street, Suite 300,
Boulder, CO 80302, USA.
2
NASA Headquarters, Washington, DC
20546, USA.
3
Applied Physics Laboratory, Johns Hopkins Uni-
versity, 11100 Johns Hopkins Road, Laurel, MD 20723, USA.
4
NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA.
5
Southwest Research Institute, Post Office Drawer 28510, San
Antonio, TX 78228, USA.
6
NASA Ames Research Center, Moffett
Field, CA 94035, USA.
7
Jet Propulsion Laboratory, 4800 Oak Grove
Drive, Pasadena, CA 91109, USA.
8
Lunar and Planetary Laboratory,
University of Arizona, Tucson, AZ 85721, USA.
9
Search for Extra-
terrestrial Intelligence (SETI) Institute, 515 North Whisman Road,
Mountain View, CA 94043, USA.
10
Lunar and Planetary Institute,
3600 Bay Area Boulevard, Houston, TX 77058, USA.
11
European
Southern Observatory, Casilla 19001, Santiago 19, Chile.
*To whom correspondence should be addressed. E-mail:
spencer@boulder.swri.edu
360 330 300 270 240 210 180 150 120 90 60 30 0
W. Longitude
−90
−60
−30
0
30
60
90
Latitude
A
Masubi
N. Lerna
Tvashtar
Zal
Amirani
Thor
Prometheus
Zamama
Marduk
Kurdalagon
B
α = 49
Masubi, Galileo
C
Masubi, NH
α = 31
D
Masubi, NH
α = 83
E
N. Lerna, Voyager
F
N. Lerna, NH
Fig. 1. (A) Global map of Io derived from eight LORRI images obtained at
phase angles between 26° and 47°, showing volcanic activity detected by NH.
See fig. S1 for an unannotated version. Yellow ovals denote areas with new,
faded, or shifted plume or other volatile deposits since the last Galileo images
in 2001. Green circles denote areas where probable new lava flows have
occurred. Cyan diamonds indicate locations of active plumes (table S1), and
orange hexagons are volcanic hot spots detected by LEISA. For plumes and hot
spots, symbol size indicates the approximate relative size and brightness of the
features. (B to F) Comparison of NH LORRI and earlier images (7)ofmajor
surface changes at Masubi (45°S, 57°W) and North Lerna (55°S, 290°W),
reprojected to a consistent geometry. The scale bars are 200 km long, and a is
the solar phase angle. At Masubi, old lava flows seen by Voyager and Galileo
(B) have been obscured at low phase angles (C) by plume deposits associated
with what is probably a new dark lava flow. The old flows are still seen by NH
at high phase angles (D), suggesting the plume deposits are not thick enough
to obscure the surface texture of the old flows. At North Lerna, a recent
eruption has generated a 130-km-long dark feature (F), probably a lava flow,
as well as an active plume that has produced a concentric pattern of deposits.
www.sciencemag.org SCIENCE VOL 318 12 OCTOBER 2007
241
SPECIALSECTION
on January 26, 2009 www.sciencemag.orgDownloaded from

(9) except for a bright new hot spot that we call “East
Girru” at 22°N, 235°W , 130 km east of the known
volcano Girru. This hot spot location corresponds to
an inconspicuous dark linear feature, possibly an old
fissure eruption, in Galileo images. No associated
albedo change is visible in sunlit LORRI images;
perhaps East Girru is a very young eruption that has
not had time to produce observable albedo cha nges.
No plume was seen in reflected light at East Girru,
but a detached glow 330 km directly abov e the hot
spot was seen in eclipse images (Fig. 2A), suggesting
some associated gas output.
LORRI eclipse images show numerous faint point
sources of emission (Fig. 2A), particularly near the
sub-jovian and anti-jovian points (on the equator at
longitudes 0° and 180°W), with a typical brightness of
~100 kRayleigh assuming a 15-km-by-15-km source
region. These were also seen in Galileo eclipse images
(6). These spots all correspond to low-albedo volcanic
centers (Fig. 2D), but because simultaneous LEISA
images show no corresponding cluster of bright spots
in the near-infrared (Fig. 2, E and F), where volcanic
thermal emission dominates, it is likely that a non-
thermal mechanism, probably plasma-related, creates
the sub-jovian and anti-jovian clusters of point-source
emission. Most of these spots are less than about 30
km in size, suggesting a near-surface origin.
A fortuitous major eruption of the Tvashtar vol-
cano during the NH flyby provides a comprehen-
sive view of a large, sulfur- rich Pele-class (3)
volcanic plume on Io. Tvashtar, a series of calderas
centered near 62°N, 122°W, has been one of Io’s
most active volcanos in recent years. An active
period from late 1999 to early 2001 (10–12 )
produced a large infrared hot spot, plume, and
orange pyroclastic deposits and was followed by
quiescent conditions seen in late 2001 (1, 13)and
early 2003 and 2004. Renewed thermal emission in
April to May 2006 (14) may have been an earlier
phase of the 2007 eruption seen by NH. Continued
thermal emission from Tvashtar was seen by
ground-based observations during and after the NH
encounter from 18 January (15) to 27 May 2007.
The 2007 Tvashtar plume was first seen in back-
scattered light in 260-nm wavelength images from
the Hubble Space T elescope (HST) on 14 February
2007 (Fig. 3A) and again in absorption in Jupiter
transitimageson21February(Fig.3B).Absorption
in the 260-nm wavelength region suggests, by
analogy with previous HST observations of the Pele
plume, that the Tvashtar plume is rich in S
2
gas (16),
as also inferred from the orange color of its plume
deposits seen previously by Galileo (12, 17).
NH imaged the Tvashtar plume on 39 occasions
over 7.8 days, at phase angles between 7° and 159°
and LORRI resolutions between 12 and 38 km per
pixel. The plume height was remarkably constant,
varying between roughly 320 and 360 km, and full
width was about 1100 km, consistent with the diam-
eter of the pyroclastic deposits (Fig. 1). The plume
had a bright top in all images (Fig. 3, C, D, and F to
J), very similar to Voyager images of the Pele plume:
This morphology is not consistent with simple bal-
listic models of plume particle flight, as noted for Pele
(18), but is consistent with hydrodynamic models
with entrained particles that include a gas shock front
at the top of the plume (19). Most Tvashtar plume
images show little evidence for a central upgoing
column of particles (e.g., Fig. 3C), suggesting that the
observed particles may condense out of the plume
rather than being directly ejected from the vent.
The plume contains remarkable time-variable
filamentary structures similar to those glimpsed in
the single high-resolution Voyager 1 image of the
Pele plume. This structure allows tracing of motion
within the plume in a sequence of five images of the
upper part of the plume obtained at 2-min intervals
on 1 March (Fig. 3, F to J, and movie S1). Speeds
projected on the plane of the sky are 0.4 to 0.7 km s
−1
(Fig. 3E), comparable to expected ballistic ejection
speeds for a 350-km-high plume (~1.0 km s
−1
), and
accelerate as plume features fall toward the surface.
Features appear to slide down the upper surface of
the plume rather than tracing ballistic trajectories
originating at the vent.
The source of the Tvashtar plume is associated
with by far the brightest hot spot seen by NH (Fig. 4).
A B
0
30
240
270
300
330
P
EG
C D
E
0
30
240
270
300
330
F
Fig. 2. Images of Io in Jupiter eclipse. (A) LORRI image taken at 27 February 14:37 UT with an effective
exposure time of 16 s. Dark blotches and straight lines are artifacts. The brightest spots (P, Pele; EG, East Girru)
are thermal emission from active volcanoes, and more diffuse emission is from the plumes and atmosphere (6).
(B) Same image with latitude/longitude grid showing glowing plumes (plume sources, table S1, indicated by
red diamonds). (C) Simulated sunlit view with the same geometry, based on sunlit LORRI images (Fig. 1A). (D)
Combined sunlit (cyan) and eclipse (red) image, showing that all pointlike sources of emission in the eclipse
image correspond to low-albedo volcanic centers. (E)A~2.3-mm LEISA eclipse image at 27 February 15:31 UT,
showing thermal emission from active volcanoes. Elongation of the hot spots is an artifact. (F) Combined visible
albedo (cyan) and LEISA thermal emission (red) image, showing the sources of the volcanic emission.
12 OCTOBER 2007 VOL 318 SCIENCE www.sciencemag.org
242
New Horizons at Jupiter
on January 26, 2009 www.sciencemag.orgDownloaded from

Thermal emission was observed on multiple occa-
sions by LORRI, LEISA, and by MVIC at wave-
lengths from 2.5 to below 0.7 mm. The hot spot
location, 62.5°N, 122.5°W , coincides with the fire
fountains seen at Tvashtar by Galileo in November
1999. The spectrum can be fit with a single tem-
perature blackbody at 1287 K from 1.25 to 2.04 mm,
providing a lower limit to the magma temperature,
comparable to Galileo estimates (12). Assuming a
temperature of 1200 K for the Tvashtar hotspot, an
area of 49 km
2
is derived from the brightness in
LORRI images, comparable to the ~25 km
2
area of
the incandescent fire fountain seen by Galileo at
Tvashtar in November 1999 (12). The isothermal
blackbody emission spectrum at close to magmatic
temperatures is also consistent with an energetic erup-
tion such as a fire fountain, rather than, for instance,
spreading and cooling lava flows (20, 21). T emper-
atures are consistent with basaltic lava composi-
tion: Exotic high temperature magmas, inferred from
some Galileo observations (22), are not required
either at Tvashtar or other hot spots seen by NH.
References and Notes
1. E. P. Turtle et al., Icarus 169, 3 (2004).
2. K. D. Retherford et al., Science 318, 237 (2007).
3. A. S. McEwen, L. Soderblom, Icarus 55, 191 (1983).
4. P. Geissler, D. Goldstein, in Io After Galileo,R.M.C.Lopes,
J. R. Spencer, Eds. (Praxis, Chichester, UK, 2007), pp. 163–192.
5. “Pele-type” plumes are thought to result from direct
ejection of gas from a volcanic vent, whereas the
smaller “Prometheus-type” plumes may result from
remobilization of surface volatiles by lava flows.
6. P. E. Geissler et al., J. Geophys. Res. 106, 26137 (2001).
7. U.S. Geological Survey Astrogeology Research Program,
http://astrogeology.usgs.gov/Projects/JupiterSatellites/io.html.
8. P. E. Geissler et al., Icarus 172, 127 (2004).
9. R. M. C. Lopes, J. Radebaugh, M. Meiner, J. Perry,
F. Marchis, in Io After Galileo,R.M.C.Lopes,J.R.Spencer,
Eds. (Praxis, Chichester, UK, 2007), pp. 307–323.
10. R. R. Howell et al., J. Geophys. Res. 106, 33129 (2001).
11. F. Marchis et al., Icarus 160, 124 (2002).
12. M. P. Milazzo et al., Icarus 179, 235 (2005).
13. F. Marchis et al., Icarus 176, 96 (2005).
14. C. Laver, I. de Pater, F. Marchis, Bull. Am. Astron. Soc.
38, 612 (2006).
15. J. A. Rathbun , J. R. Spencer, paper presented at the 39th
annual American Astronomical Society, Division for
Planetary Science meeting, Orlando, FL, 9 October 2007.
16. J. R. Spencer, K. L. Jessup, M. A. McGrath, G. E. Ballester,
R. Yelle, Science 288, 1208 (2000 ).
17. A. S. McEwen et al ., Icarus 135, 181 (1998).
18. R. G. Strom, N. M. Schneider, in Satellites of Jupiter,
D. Morrison, Ed. (Univ. of Ariz ona Press, Tucson, AZ,
1982), pp. 598–633.
19. J. Zhang et al., Icarus 172, 479 (2004).
20. J. A. Stansberry, J. R. Spencer, R. R. Howell, D. Vakil,
Geophys. Res. Lett. 24, 2455 (1997).
21. A. G. Davies et al., J. Geophys. Res. 106, 33079 (2001).
22. A. S. McEwen et al ., Science 281, 87 (1998).
23. We thank the entire NH mission team, particularly D. Rose
and E. Birath, and our colleagues on the NH science team.
NH and the ancillary investigations described here are funded
by NASA, whose financial support we gratefully acknowledge.
Supporting Online Material
www.sciencemag.org/cgi/content/full/318/5848/240/DC1
Fig. S1
Table S1
Movie S1
10 July 2007; accepted 19 September 2007
10.1126/science.1147621
Fig. 3. The Tvashtar plume.
(A) Discovery image by HST
in backscattered light in the
F255W filter (central wave-
length = 260 nm). The red
diamond indicates the plume
source. (B) HST image of
260 nm absorption by the plume
against Jupiter: 260 nm (blue)
plus 330 nm (green) plus
410 nm (red) color compos-
ite. Other images are in visi-
ble light from NH LORRI. The
scale bar is 200 km long, and
the yellow star indicates the
projected location of the hot
spot at the plume source. The
dashed line is the terminator.
(C) Highest-resolution view of
the full plume, at a resolution
of 12.4 km per pixel and
phase angle of 102°, show-
ing the filamentary structure.
Images are sharpened by
unsharp masking: the dark
line at the edge of the disk is
an artifact of the sharpening.
(D) Image at 145° phase
angle at 22.4 km per pixel,
showing the time variability
of the details of the plume
structureandthepersistent
bright top. (F to J)Sequence
of frames at 2-min intervals
showing dynamics in the up-
perpartoftheplume(the
source is on the far side of
Io). Colored diamonds track
individual features whose
speeds, projected on the plane of the sky, are shown in (E).
F 03/01 23:50:31
G 03/01 23:52:31
H 03/01 23:54:31
I 03/01 23:56:31
J 03/01 23:58:31
C 02/28 11:04:22
D 03/02 06:07:22
E Plume speeds
0 2 4 6 8
Time, Minutes
0.
3
0.
4
0.
5
0.
6
0.
7
0.
8
Projected Speed, km/sec
A HST, 02/14 B HST, 02/21
Fig. 4. LEISA spectra of volcanic thermal emission from Tvashtar, Pele, and East Girru, with best-fit
blackbody curves superposed. Vertical axis units are GW steradian
–1
mm
–1
.
www.sciencemag.org SCIENCE VOL 318 12 OCTOBER 2007
243
SPECIALSECTION
on January 26, 2009 www.sciencemag.orgDownloaded from