EGAD

Deep Impact (Live): Much Cooler Than a Mediocre Blockbuster

This image from NASA TV is a view from Deep Impact's flyby showing the impactor colliding with comet Tempel 1.

Image credit: NASA/JPL-Caltech/UMD

Nursing my climate-change-induced sniffles. Evidently the ol' constitution needs some time to adjust to this odd phenomenon of water issuing forth from the sky. No worries; as there have been no ScavHunts or final exams recently, escalation to a bout of Hacking Death Plague is unlikely. Now, where better to recouperate than in my lab? After all, it's climate-controlled, has a limitless supply of (effectively) free tissues, and a right speedy network link. This last feature is just what I needed to keep an eye on tonight's big feature, anyway.

Deep Impact should not be confused with the mediocre Summer-of-the-Killer-Comet movie which nevertheless managed to be head and shoulders better than its Asteroid Flick With Animal Crackers sibling. No, today I'm watching as the Deep Impact space probe blasts an artificial crater in comet Tempel 1 to dig up a bit of the stuff the Solar System was originally made from.

(Okay, so movie releases in 1998. Mission is approved in 1999. There may have been some inspiration in the naming.)

In addition to the JPL site, NASA has a site as well. They have mostly similar information, and they both use the same "Near Real Time" image feed, which is probably the best way for those of you using modems to watch the action. At this point the impactor -- a desk-sized lump of copper and gadgetry -- has been released towards the comet, and the host spacecraft has dodged (barely) out of the path of the nucleus. Live NASA TV coverage from JPL mission control just began, for those of you on broadband. Next, it's about an hour until the impactor's onboard nav computer will start making final course corrections to ensure it hits the nucleus square on, and the actual collision takes place at 12:52 AM CDT (+/- 17 seconds, they're saying now).

Clearly, an opportunity like this doesn't waltz along every day. Besides the sensors on the Deep Impact probe and the impactor, then, astronomers the world over are getting in on the act, pointing most anything with a camera at the event. At about 10th magnitude and at about the distance of Mars's orbit, you'd need pretty dark skies to see this comet under normal conditions. But since nobody knows how much material will be excavated by the impact, it's also an open question as to what will be visible from the Earth. Sadly, though, nothing of great interest will be seen from North America, as it'll have set by the time of impact. Hawaii will have the best views from the ground.

Running updates in the post body.

[Update 22:35 - 22:47 CDT]

It's reported that the impact probe has 7.27 kilograms of fuel remaining for automatic navigation, which is good to change its course by up to 32 meters per second, or roughly 71 mph for the Americans. As the last course correction called for around 1 m/s of correction, this should be plenty.

There seems to be a problem with the high-gain antenna, which is worrying the engineers, since we're 14 minutes out from putting the probe on auto-pilot. Sounds like a pointing problem has cropped up since they switched from one receiving station to another that is corrupting data. This in turn is thought to be connected to motion of the filter wheel -- a mechanism that lets the onboard cameras select which wavelengths of light to observe at -- putting an unexpected torque on the spacecraft. But they're checking into other possibilities, too.

Some commands have been sent, and in roughly 7 minutes the next downlink will tell whether data is coming down intact.

[Updates 23:21 - 23:38 CDT]

After switching back to the Goldstone receiving station from Canberra, data seems to be coming down smoothly now, although I still hear some background chatter among the engineers trying to diagnose the filter wheel anomaly. This is creating some noise in the spacecraft alignment data, which would complicate fine navigation.

In other navigational news, word just came through that the impactor has successfully computed and executed its first course correction. It was a very small one, as so far the targetting has been very good. But it's only recently that the impactor got close enough to resolve the cometary nucleus in the targetting sensor camera, which allows it to figure out where the center is and aim for it. On the other hand, it sounds like the auto-nav might be a little too independent-minded, The impactor computer has been sporadically rejecting commands.

The NASA TV feed is getting congested, but as long as the audio keeps coming through I should be able to continue piecing together what's going on.

[Update: 4 July 00:20 - 00:35 CDT]

Applause in Mission Control as word comes in that ITM2 (the second impactor course correction) went off successfully. About 6.7 kg of fuel left, and the impactor is evidently within under a kilometer from where the ground solution said it should be. Which is to say, the astrometrics folks who did the preliminary work from Earthside were right on the ball.

Although the latest images haven't shown up on the web yet, pictures of a recognizably cratered eggplant have been flashed up on screen in MC, fresh from the hi-res camera by way of the science team's laptops. The returned data has already surpassed the Giotto flyby of Halley and the DS1 Borreley images in resolution. At least, after some fairly aggressive image deconvolution. But that's allowed, since I'd assume the point-spread functions of these cameras are very well known.

Five minutes to the final impactor rocket burn. Fifteen minutes to impact.

[Update: 00:47 CDT]

Okay, so this probably isn't entirely kosher, but here's a taste of what we're seeing on screen now.

As the impactor heads into the debris field, the flight engineers have shut down all of its fault tolerance systems; it's now set to forge on ahead no matter what happens to it. Which is the point -- even if it is damaged by ejecta on the way in, we'd still like to get any data it can still give us.

[Update: 01:15 CDT]

Impact! Now that the impactor craft has been destroyed, and the host probe is in shield mode to protect its instruments from the dust tail, they're playing back images. Here's what I saw:

The impactor is "performing quite well for a spacecraft that's about to be vaporized," remarks the mission director.

We have impact, says the telemetry team! Looks like the impactor camera didn't quite make it all the way down, though. Still, good enough to provoke dancing in Mission Control. After a few minutes of that, the Announcer comes on. "Attention everyone. We still have one more spacecraft to worry about, so let's settle down."

Flipping back to the flyby craft, we have dramatic visual evidence that impact has taken place, and some closeups of the surface during the close approach.

[Update: 01:32 CDT]

Now that we're in Shield Mode, they've turned the fault control software back on, and begun downlinking the rest of the flyby data.

Nope, that's not an artist rendering; this is what the flyby probe actually saw. Actually, there seems to be rather more energy in that eruption than the impactor could have delivered. This suggests that comets are even more volatile than was generally thought, and possibly even under pressure internally, which has the scientists at the JPL extremely excited.

And on that note, I'll call it a night. Hopefully by tomorrow there will be tasty fully-reduced imagery to drool over. Although I'm actually more interested in the ejecta spectroscopy, as that'll tell us what was actually in there.

[Update: 4 July 2005]

Most of the images above are now available at considerably higher quality from the JPL and NASA mission websites. Particularly cool are the movies (medium-res flyby, hi-res flyby, and impactor targeting camera views) stitched together from images returned near the time of impact. But I'll leave the above pictures up, anyway, to retain the feel of watching things unfold in the moment.

The correct image credit for the screencaps above is, I believe, "NASA/JPL-Caltech/UMD".