NOTE: Click on the image to view it at its highest resolution.
This image shows the direct impact of Shoemaker-Levy fragment N into Jupiter. These data were recorded by the camera (SSI) aboard the Galileo spacecraft, which is currently en route to Jupiter. This gray-scale image is shown in false-color to enhance visibility of the impact flash.
The entire picture covers the duration 10:28:13 UT to 10:30:40 UT on 1994 July 20 (times are those at which an earth-based observer would have received the light from Jupiter). The bright swaths are the planet Jupiter, scanned across the CCD detector. The N impact is seen just off the left (terminator) side of the central swath.
This "drift-scan" observation was made using the near-infrared 889-nm (methane band) filter, which reduces the brightness of Jupiter relative to the impact flash. This is effectively a multiple-exposure, intentionally smeared picture of Jupiter. The same filter and observing mode were used for fragment K, so the two data sets can be compared directly. This mode serves to increase sensitivity, time coverage, and time sampling; it also permits data to be returned more efficiently than for other imaging modes with Galileo's small, back-up antenna. Jupiter was allowed to drift across the detector for about 25 seconds before the camera was repositioned for the next swath. Altogether there are five separate 25-sec long swaths of Jupiter; they were laid down at an angle, so that any impact events would not be overlapped by Jupiter itself.
The horizontal stripes appear because not all of the data will be returned to Earth. To allocate downlink time for playback of data from other events and other instruments aboard Galileo, only about 4 out of every 12 lines were relayed to Earth. However, the gaps are no longer than about 0.6 seconds, and within each group of lines the time resolution is 0.072 sec.
There is some confusion in the search for impact events in these N data due to the presence of trails of Jupiter's Galilean satellites (the very faint, blue trails seen between and to the sides of the Jupiter swaths). By chance, the trails of satellites Io and Europa (caused by multiple exposure from adjacent swaths) lie in nearly the same place as a trail from any potential impact event. However, the brightness of the satellite trails is much less than that of the N impact and does not significantly affect the data quality. Furthermore, the satellite trails (and Jupiter trails) follow a well-defined pattern from the periodically repeated jitter of the spacecraft's scan platform.
At the time of the observations, the Galileo spacecraft was about 1.5 astronomical units (about 150 million miles) from Jupiter on its way to a rendezvous with that planet in December 1995. The camera resolution was about 2500 km/pixel. From Galileo's viewpoint, the impact site was on the dark side of Jupiter, but directly visible to the camera. For observers on earth, the impact site was just beyond limb, on the half of Jupiter we could not see, so Galileo has provided the only direct pictures of the comet impacts.
At its peak brightness, the N impact was about half as bright as the peak of the K impact, or about 4% of the total brightness of Jupiter (at this phase angle of 51 deg and in this methane-band filter). The detected flash begins at about 10:29:17 UT and lasts for about 14 seconds. The lightcurve of the event rises very rapidly in the first couple seconds, as in other events detected by Galileo; but, unlike the others, it does not extend very long in time. Our preliminary interpretation is that we are seeing primarily the bolide (meteor) phase of the event and little of the ensuing fireball, probably because this event was less energetic than the others we have observed. The observed time of impact is within error bars of the estimate of impact time derived from observations of the N fragment's small scar by HST and the best guess of P. Chodas and D. Yeomans based on systematics of the other SL9 phenomena.
The Galileo/SSI experiment took data on 4 of the impact events: K, N, V, and W. Data for event V were taken in a less sensitive observing mode; given that V turned out to be a very weak event as observed from Earth, it was decided not to allocate resources to return the V data, but to concentrate on K, N, and W. Although several images returned from the W sequence already show the bolide flash, most of the W images will not be returned until late December. After that, preparations must begin for the Jupiter encounter.
For further descriptions of the experiment, see the diagrams and captions that accompany the images of the K and W events.
Clark Chapman and Bill Merline,
on behalf of Mike Belton, Torrence Johnson, Ken Klaasen,
Catherine Heffernan, and the Galileo Imaging Team.
21 November 1994
Images, Images, Images