formed by gradual sapping as liquefied SO 2 seeps out at the base (36 ), perhaps exploiting and enlarging preexisting joints and fractures. The height of the mesa forming the eastern margin of Tvashtar Catena (Fig. 7) is 1 km, the depth at which SO 2 is expected to become liquid (30). Diffuse white patches often appear to emanate from the bases of scarps on Io, consistent with plumes of SO 2 expected to form when the liquid reaches the triple point near Io’s surface. Discussion. The dominant eruption styles on Io may vary with latitude. At low lati- tudes, we see many long-lived eruptions with insulated flow fields and often associated with Prometheus-type plumes, as well as short-lived eruptive episodes (Pillan) and lava lakes (Pele). At high latitudes, the erup- tions are mostly short-lived but high-volume outpourings of lava, probably with lava foun- tains such as that at Tvashtar (27 ). We have never seen an active plume at high latitudes, but we do see new color patterns indicative of short-lived plumes. One interpretation is that the lithosphere is thicker at high latitudes, such that only large batches of magma are able to ascend to the surface. A subject of great interest for understanding global change is whether terrestrial flood lavas have been emplaced rapidly in open channels or sheet flows or relatively slowly through insu- lated (crusted-over) tubes or sheet flows (37 ). Most terrestrial flood lavas are highly eroded, so the emplacement style is contentious. On Io, we see examples of both rapidly emplaced flows (Pillan and Tvashtar) and flows emplaced over many years or decades (Zamama, Prometheus, Amirani, and Culann). These ac- tive flow fields provide important clues to the emplacement of ancient flood lavas on Earth and other planets. The formation and destruc- tion of landforms such as mountains and calderas are also much more rapid on Io than on other planets, so Io is a unique laboratory to study processes normally inferred from the incomplete geologic record. References and Notes 1. W. J. O’Neil et al., in The Flight of Project Galileo as Reported Annually to the IAF/AIAA, IAF-96-Q.2.01, 1 (International Astronautical Federation, Paris, 1997). 2. The majority of the I24 images were acquired in a special mode (2  2 pixel summation and a fast 2.6-s readout time) designed to minimize radiation noise. We expected the radiation noise to be severe in images acquired close to Io, but the image quality proved much better than expected. Unfortunately, the summation mode, which worked correctly through orbit C21, produced garbled images in I24. Many of the images were reconstructed with an innovative algorithm devised at the Jet Propul- sion Laboratory ( JPL) with the LabVIEW software from National Instruments (Austin, TX). However, the pho- tometry remains severely compromised, eliminating useful color data, and parts of some images are com- pletely unrecoverable. Full-resolution imaging modes (no pixel summing) worked correctly, but only a few partial (top one-third) frames were acquired in I24 while close to Io. I25 and subsequent encounters were replanned to acquire only full-resolution images. 3. See www.sciencemag.org/feature/data/1049308.shl for a table of I24 and I25 image characteristics and the Web figures. 4. A. S. McEwen et al., Icarus 135, 181 (1998). 5. R. Lopes-Gautier et al., Icarus 140, 243 (1999). 6. P. E. Geissler et al., Icarus 140, 265 (1999). 7. P. E. Geissler et al., Science 285, 870 (1999). 8. M. H. Carr et al., Icarus 135, 146 (1998). 9. A. S. McEwen et al., Science 281, 87 (1998). 10. T. V. Johnson and L. A. Soderblom, in Satellites of Jupiter, D. Morrison, Ed. (Univ. of Arizona Press, Tuc- son, AZ, 1982), pp. 634 – 646. 11. L. Keszthelyi and A. S. McEwen, Icarus 130, 437 (1997); A. S. McEwen, R. Lopes-Gautier, L. Keszthelyi, S. W. Kieffer, in Environmental Effects on Volcanic Eruptions: From Deep Oceans to Deep Space, J. R. Zimbelman and T. K. P. Gregg, Eds. (Plenum, New York, in press). 12. Plumes have given names only (such as “Pele”), whereas surface features are called a patera (irregular or complex depression), fluctus (flow), mons (mountain), mensa (mesa), planum (plateau), or catena (chain). Most vol- canic centers include several features, so in prac- tice we often drop the feature name and refer to the entire complex by its given name. 13. J. R. Spencer et al., Geophys. Res. Lett. 24, 2471 (1997). 14. J. R. Spencer, K. L. Jessup, M. A. McGrath, G. E. Ballester, R. Yelle, Science 288, 1208 (2000). 15. A. G. Davies et al., Lunar Planet. Sci. XXX (1999) [CD-ROM]. 16. L. Keszthelyi and A. S. McEwen, Geophys. Res. Lett. 24, 2463 (1997). 17. C. B. Phillips, dissertation, University of Arizona, Tuc- son, AZ (2000). 18. L. Keszthelyi, A. S. McEwen, Th. Thordarson, J. Geo- phys. Res., in press. 19. J. R. Spencer et al., Science 288, 1198 (2000). 20. S. Thorarinsson, Bull. Volcanol. 2, 1 (1953). 21. D. A. Williams, A. H. Wilson, and R. Greeley, J. Geo- phys. Res. 105, 1671 (2000); D. A. Williams et al., Eos (spring meet. suppl.), in press. 22. K. Hon et al., Geol. Soc. Am. Bull. 106, 251 (1994); B. C. Bruno, G. J. Taylor, S. K. Rowland, P. G. Lucey, S. Self, Geophys. Res. Lett. 19, 305 (1992). 23. T. N. Mattox, C. Heliker, J. Kauahikaua, K. Hon, Bull. Volcanol. 55, 407 (1993). 24. R. Lopes-Gautier et al., Science 288, 1201 (2000). 25. S. W. Kieffer et al., Science 288, 1204 (2000). 26. D. S. Acton, M. E. Brown, B. F. Lane, in preparation. 27. J. A. Stansberry, J. R. Spencer, R. R. Howell, C. Dumas, D. Vakil, Geophys. Res. Lett. 24, 2455 (1997). 28. S. A. Fagents, D. A. Williams, R. Greeley, Eos (fall meet. suppl.) 80 (no. 46), F625 (1999). 29. L. Wilson and J. W. Head, Nature 302, 663 (1983); J. W. Head and L. Wilson, Lunar Planet. Sci. XXXI (2000) [CD-ROM]. 30. S. W. Kieffer, in Satellites of Jupiter, D. Morrison, Ed. (Univ. of Arizona Press, Tucson, AZ, 1982), pp. 647–723. 31. S. E. Heslop, L. Wilson, H. Pinkerton, J. W. Head, Bull. Volcanol. 51, 415 (1989). 32. J. S. Kargel, P. Delmelle, D. B. Nash, Icarus 142, 249 (1999). 33. P. M. Schenk and M. H. Bulmer, Science 279, 1514 (1998). 34. The compressive stress induced by globally uniform subsidence exceeds 1 kbar at a depth of 2.5 km. For comparison, the stresses induced by tides are expect- ed to be between 2 and 6 bar. 35. J. M. Moore, R. J. Sullivan, R. T. Pappalardo, E. P. Turtle, Lunar Planet Sci. XXXI (2000) [CD-ROM]. 36. J. F. McCauley, B. A. Smith, L. A. Soderblom, Nature 280, 736 (1979). 37. S. Self, Th. Thordarson, L. Keszthelyi, in Large Igneous Provinces: Continental, Oceanic, and Planetary Flood Volcanism, J. J. Mahoney and M. F. Coffin, Eds. (American Geophysical Union, Washington, DC, 1997), pp. 381– 410. 38. We thank J. Erickson and the Galileo team at JPL for their spacecraft recovery efforts during I24 and I25 and G. Levanus and G. Wells of JPL for the I24 image reconstruction. 8 February 2000; accepted 18 April 2000 Io’s Thermal Emission from the Galileo Photopolarimeter- Radiometer John R. Spencer, 1 * Julie A. Rathbun, 1 Larry D. Travis, 2 Leslie K. Tamppari, 3 Laura Barnard, 3 Terry Z. Martin, 3 Alfred S. McEwen 4 Galileo’s photopolarimeter-radiometer instrument mapped Io’s thermal emission during the I24, I25, and I27 flybys with a spatial resolution of 2.2 to 300 kilometers. Mapping of Loki in I24 shows uniform temperatures for most of Loki Patera and high temperatures in the southwest corner, probably resulting from an eruption that began 1 month before the observation. Most of Loki Patera was resurfaced before I27. Pele’s caldera floor has a low temperature of 160 kelvin, whereas flows at Pillan and Zamama have temperatures of up to 200 kelvin. Global maps of nighttime temperatures provide a means for estimating global heat flow. The photopolarimeter-radiometer (PPR) is a simple aperture photometer on the Galileo scan platform, with a selection of broadband infrared filters including a wide-open filter that is sen- sitive to light over the full visible to 100-m sensitivity range of the detector (1). During the close flybys (2), PPR obtained both dedicated raster scans of Io and “ride-along” sequences of data obtained simultaneously with the near- infrared mapping spectrometer (NIMS) or sol- id-state imaging (SSI) instruments. Calibration is with reference to dark sky and an onboard calibration source, and we estimate brightness temperatures (T B ’s) to be accurate to within 10 K at 250 K and 5 K at 125 K (3). 1 Lowell Observatory, 1400 West Mars Hill Road, Flag- staff, AZ 86001, USA. 2 Institute for Space Studies, NASA–Goddard Space Flight Center, 2880 Broadway, New York, NY 10025, USA. 3 Jet Propulsion Laborato- ry, 4800 Oak Grove Drive, Pasadena, CA 91109, USA. 4 Lunar and Planetary Laboratory, University of Arizo- na, Tucson, AZ 85721, USA. *To whom correspondence should be addressed. E- mail: spencer@lowell.edu G ALILEO :I O U P C LOSE 19 MAY 2000 VOL 288 SCIENCE www.sciencemag.org 1198