Editorials

Go the Extra Mile, NASA, And Fund Another Deep Impact Mission

Deep Impact needs a trajectory change soon if it is to have enough propellant for a new mission. That change remained forbidden by NASA Headquarters as the spacecraft team celebrated their hard-won success on July 4, a sign of money troubles and uncertainty about directions at the agency. Clear heads prevailed on July 5 and the course change was approved--but not the extended mission. We urge NASA to go all the way and support this very cost-effective proposal that should make fresh discoveries and substantially grow our knowledge of these interesting bodies, with a proven high-quality performer.

========

Cover Story

CRASH COURSE

Deep Impact scientists suspect that comet 9P/Tempel 1 has a fluffy makeup that is held together more by gravity than by the cohesiveness of its materials, based on early data from the flyby and strike of the body that kicked up more material than most imagined.

The flyby spacecraft and its instruments are in excellent shape after the close pass by the dusty comet.

Later this month, controllers will fire thrusters to target an Earth flyby in January 2008, to prepare for a not-yet-approved extended mission, possibly to the comet Boethin.

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_246_L.jpg

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_248_L.jpg

Top picture sequence (Nos. 1-7) is final frames radioed back by impactor, up until 3 sec. before it hit comet Tempel 1. The surface is darker than charcoal but 14-bit dynamic range of camera has produced good images. Pair of craters is readily visible in frame 3, and scientists believe the craft hit near the upper edge of the lower crater. At end, pointing of impactor was thrown off a fraction of a degree, probably by debris hits. Final frame 7 was taken at about 100,000 ft. while sinking at 30,000 fps. Credit: NASA.JPL-CALTECH/UMD; EDITED BY AW&ST

The science payoff highlights the comeback of Deep Impact, which at one point was almost canceled. As development progressed, the difficulty of finding and targeting a desired spot on a completely unknown body using scene recognition algorithms while closing at 22,800 mph. became apparent, and an expensive switch to a more powerful computer was required. Other difficulties included getting Jet Propulsion Laboratory autonomous navigation software to play correctly with Ball Aerospace attitude control software.

But the mission, led by principal investigator Michael A'Hearn at the University of Maryland, persevered and, with extra money and a year of schedule slip, solved the problems to achieve the July 4 success.

The mission team was most anxious about the autonomous navigation in the encounter--first, that the impactor would guide itself to a proper strike point on the comet, and second that the flyby craft would independently duplicate this targeting so it could focus its close-up lens on the right spot.

They were jubilant at the results. The impactor struck where it was supposed to, in plain view of the flyby craft, and the flyby's aimpoint was within 50 meters (165 ft.) of that spot. But a reconstruction showed that the impactor's first targeting maneuver actually pulled it off the path to the comet nucleus, though the second and third maneuvers brought it back on. That was not unexpected but still caught people's attention.

The amount and duration of material kicked out by the strike was toward the large end of expectations. On the one hand, that provides more material to examine; on the other, scientists wanted to look into the impact crater and the dust cloud obscured their view. They were using image processing late last week to try to spot the crater through the dust in closeup pictures. The comet was moving away so fast that it was back to one pixel in size in Deep Impact's most powerful telescope four days after the strike.

The University of Maryland is responsible for overall Deep Impact mission management, and project management of the Discovery-class mission is handled by NASA's Jet Propulsion Laboratory. The spacecraft was built for NASA by Ball Aerospace & Technologies Corp. in Boulder, Colo.

Deep Impact has two components--the flyby spacecraft and the impactor. The flyby's science package comprises:

*The High-Resolution Instrument (HRI). This is a powerful 30-cm. (12-in.) reflecting telescope with a 10.5-meter (34.4-ft.) focal length. It feeds both a 1,024 X 1,024-pixel CCD through a filter wheel for visual observations, and an imaging infrared spectrometer. The visual resolution was designed to be 2 microradians per pixel, or 1.4 meters/pixel at the 700-km. closest viewing distance to the nucleus. But in a repeat of the Hubble Space Telescope problem, a flaw in one of the measuring devices used to tune the HRI on the ground made it blurry so that it only resolves 7 meters/pixel. Knowing the flaw, scientists have been applying a deconvolution algorithm to the images to bring some sharpness back. All instruments are rigidly attached to the spacecraft bus, and the HRI had to be pointed accurately because the nucleus more than fills the frame near closest approach.

*The Infrared spectrometer (IR). It is very sensitive and can rapidly acquire spectra across a line of pixels. "It is the data hog," says Jessica M. Sunshine, a science co-investigator from SAIC. That's needed because the cloud of dust, ices and gases kicked up by the strike was expected to evolve rapidly, with intermediate products of chemical reactions lasting only briefly. The IR can take snapshots as fast as every 0.72 sec. when spatial and spectral coverage is reduced to 64 pixels and 512 lines, respectively. At best, it can read 512 pixels across its linear detector and resolve that into 1,024 spectral lines, but frame rate slows proportionately. It covers wavelengths from 1.05-4.8 microns, with the shorter wavelengths aimed at silicates and the longer ones aimed at gases.

An early test showed how much the state of the art had advanced. Compared with the spectrometer on the Soviet Union's Vega spacecraft, which observed Halley's Comet in 1986, Deep Impact's IR achieved roughly the same signal-to-noise ratio from 300 times farther away on a comet that was 100 times dimmer, Sunshine says. The key is a sensitive mercury-cadmium-telluride 1,024 X 512-pixel infrared focal plane array. It is passively cooled with an infrared radiator to 89K. The spatial resolution is 10 microradians/ pixel, or 7 meters/pixel at 700 km., the same as the Medium-Resolution Instrument.

*The Medium-Resolution Instrument (MRI). It is a separate 12-cm. telescope with 2.1-meter focal length, with the same type of CCD and filter wheel sensor as the HRI for visual observation only. It resolves to 10 microradians/pixel, giving a five times broader field of view than the HRI, and is used for optical navigation as well as science images.

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_257_L.jpg 

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_259_L.jpg  

Bottom sequence (frames A-H) is from medium-resolution camera on flyby craft. Its viewpoint has the images rotated about 70 deg. to the right compared with impactor sequence. Impact first appears in frame B, and in frames C and D it is so bright that pixels are saturated, probably due to incandescent gas. In frame E, dust and vapor cloud begins to emerge with umbrella-shaped leading edge. In frames F-H, sunlight from bottom is blocked by ejecta, casting a thin vertical shadow on the nucleus.

The flyby craft has a 515-kg. (1,135-lb.) dry mass and started the mission with 86 kg. of hydrazine monopropellant. It is 10.8 ft. long, 5.6 ft. wide, and 7.5 ft. high. It is three-axis stabilized using both thrusters and momentum wheels. The absolute pointing accuracy is 64 microradians (0.0036 deg.), says Monte L. Henderson, the Ball program manager. During the encounter it was allowed to dither within a several-hundred microradian deadband, or about one-tenth the HRI's 2,050-microradian field of view, says Richard Grammier, the JPL project manager.

The X-band high-gain antenna was gimballed to stay locked on Earth as the flyby craft maneuvered to keep the cameras pointed at the comet. Since there was concern that the flyby craft itself could be destroyed during the encounter, scientists wanted to get as much data down as soon as possible. The downlink rate was 200 kilobits per second. Distance from the Earth at flyby was 83 million mi., for a one-way light time of 7.4 min.

The impactor's instrument was the Impactor Targeting Sensor (ITS), which was a duplicate of the MRI without a filter wheel. After release from the flyby craft, it supplied imagery to the autonomous navigation system (AutoNav), which locked on to the comet starting as a 1-pixel target, then sought the center of brightness as it grew bigger, finally switching to a scene analysis scheme to pick the best point for flyby photography. AutoNav derives from the system pioneered on JPL's Deep Space 1, which flew past the comet Borrelly in 2001 (AW&ST Oct. 22, 2001, p. 79).

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_261_L.jpg

Flyby and impactor spacecraft were stacked at Ball Aerospace in Colorado prior to shipment to Cape Canaveral for launch. Credit: BALL AEROSPACE & TECHNOLOGIES CORP.; CALLOUTS BY AW&ST

The ITS took pictures all the way in, with the last one 3 sec. before impact. "It was at 100,000 ft., sinking at 30,000 fps., and the little box was yelling 'Pull up! Pull up!'" quipped one engineer. The impactor pictures were relayed by S-band link to the flyby craft, which forwarded them to Earth mixed with its own data.

The impactor weighed 820 lb., including 249 lb. of copper dead weight to increase the cratering effect. Copper was chosen so it would not contaminate the spectroscopy; copper is not expected to be present in the nucleus, and its spectral lines are easily removed from the data. The propulsion system had lateral divert thrusters and would roll to point them before making each of the three impactor targeting maneuvers. The craft carried 8 kg. of hydrazine, worth about 34 meters/sec. of divert velocity.

The final trajectory correction maneuver (TCM-5) of the combined spacecraft was on July 2 at 5:07 p.m. PDT Earth-received time and was a small 0.3 meters/sec. to point the pair directly at the comet nucleus. The impactor was released at 11:07 p.m. that evening, almost 24 hr. before impact; 12 min. later the flyby craft made a long 25-min. burn to slow down 120 meters/sec. and target it to a 500-km. closest approach of the nucleus, lagging the impactor.

The ITS started sending back images with the comet near center, giving the team a warm feeling. AutoNav started processing images at 8:53 p.m. on July 3, 1 hr. before impact. The first impactor targeting maneuver (ITM-1) took place at 9:21 p.m. and was 1.3 meters/sec. Post-impact analysis shows that actually increased the miss distance from the release trajectory to 7 km. but the nucleus was only subtending a few pixels in the ITS at that point and this result had been seen in simulation.

The trajectory was brought back on to the nucleus with ITM-2 making a 2.2-meters/sec. diversion at 10:17 p.m. AutoNav was still using a center-of-brightness technique and the nucleus was subtending roughly 20-40 pixels. The final ITM-3 reflected a switch in technique to scene analysis, and was a 2.3-meters/sec. burn at 10:39 p.m., just 13 min. before impact.

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_265_L.jpg

Autonomous navigation system shifted targets several times as it closed in and got better pictures. Note that first impactor targeting maneuver (ITM-1) moved the strike point 7 km. off the nucleus.

The strike occurred at 10:52 p.m., about 1.5 sec. from the expected time. The kinetic energy of the 10.2 km./sec. collision was equivalent to 4.5 tons of TNT. The impactor strike angle was 25-30 deg. above the comet horizon.

Late last week, Sunshine had examined the first 5 sec. of the voluminous IR data. It supported the idea of a comet mainly held together by gravity (AW&ST Dec. 13, 2004, p. 66). It showed two distinct characteristics--a base of 300K blackbody radiation, which is the warmed dust cloud, and broad peaks representing incandescent gases of water, carbon dioxide and organics at 1,000-2,000K, she says. That goes along with the idea that the impactor immediately created an incandescent flash lasting less than a second, and when the impactor had burrowed well under the surface the shock wave kicked out large amounts of dust--the model projected for a gravity-dominated body.

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_263_L.jpg

The flyby craft took this lookback picture 50 min. after impact with its high-resolution instrument. Dust cloud is backlit by the Sun, and fainter lines delineate the ejecta cone.

The spectrometer slit, which was more than 5 km. wide at the comet, was pointed off the crater by 1-2 km. and missed getting crater measurements, but this may have improved the gathering of impact data.

The gravity theme was also supported by lookback photos of the ejecta cone, showing it to be connected to the surface and to be well formed. Were the cone detached, it would be a sign of a comet held together more by material cohesiveness.

"Tempel 1 is the first comet encountered that has what are undeniably impact craters," says Sunshine. The shape of surface features is a battle between weak materials and weak gravity.

More than 50 telescopes around the world were part of the impact observing program, as well as Hubble, the European Space Agency's Rosetta, and other spacecraft. Hubble saw Tempel 1's brightness increase 5-6-fold owing to the impact.

The flyby craft had an identical AutoNav program on board to aim at the same spot the impactor was finding independently. The difficulty was the flyby craft was looking at the comet at up to a roughly 45-deg. different angle. An algorithm warped the MRI image to the best approximation of what the impactor was seeing before feeding it to AutoNav, but this scheme could have been upset by a large bump or depression on the comet. In the event, AutoNav on flyby craft and impactor produced remarkably similar targets, and scientists are happy with them.

http://www.aviationnow.com/media/images/awst_images/large/AW_07_11_2005_266_L.jpg

The impactor took this photo 5 min. before strike. Scientists are puzzled by the varied topography, including smooth patches, one to the left below the beltline and the other at the 1-o'clock position from the large crater above the center.

The flyby craft took pictures for 13 min. past impact, until 11:05 p.m., when it turned to a shield attitude for protection against any dust particles at closest approach. At 11:51 p.m., it turned back to resume imaging the comet and continued doing so through July 6; then the science instruments will undergo a calibration sequence. They appear to be undamaged, says Grammier.

The thruster firing to aim Deep Impact at Earth for a possible extended mission will probably be accomplished by July 20-24, he says. When the comet was hit, the thruster firing was forbidden by NASA headquarters, but the Solar System Div. gave permission on July 5.