The NASA Deep Impact mission consisted in two spacecrafts: an impactor, weighting 364 kg; and a flyby spacecraft for observing the impact and relaying data from the impactor. The main goal of the mission was to study the interior and outer layers of a comet. Until the impact, very little was known of the internal structure and the physical evolution of the outer layers of a comet nucleus. Most of what we know relies primarily on theoretical models. The relationship between the coma’s composition and the nucleus composition is also uncertain. Even if the coma is formed by material from the nucleus, there are several physical and chemical processes that rapidly affect the material ejected from the nucleus.

Comets are remnants of the early stages of the formation of our Solar System and thus contain the most pristine material from that era, as well as clues to its subsequent evolution. Whatever evidence we have into their internal composition comes either from remote observation and modelling of the dust and gases that are lifted off the surface, or from in-situ analysis of data from recent spacecraft flybys. Deep Impact was designed to provide a first look at the interior of a comet by striking the surface to expose the material underneath the opaque crust.

The target comet was comet 9P/Tempel1. This is one of a class of comets known as the Jupiter-family of comets, most of which are believed to have formed in the trans-Neptunian region. These objects have low inclination orbits and typically take less than 20 years to orbit the Sun. Their orbits are strongly influenced by Jupiter, hence their name. 9P/Tempel 1 orbits the sun once every 5.5 years, and the Deep Impact encounter was scheduled to take place at perihelion, when the comet was at 1.5 and 0.9 Astronomical Units from the Sun and Earth, respectively.

Deep Impact was designed so that much of the mission-critical science would be done from Earth-based telescopes. These facilities would observe the comet before, during, and after impact. This was an unprecedented coordinated observational campaign, which included over 550 whole or partial nights of observation using 73 ground-based telescopes at 35 observatories. These facilities would observe the comet’s evolution in wavelength regimes and timescales inaccessible to the spacecraft (The Tempel1 Observing Collaborators Team).

The Roque de Los Muchachos Observatory played a substantial role in this campaign. Observations started in 2000. But the interesting part of the game started on July 2nd, 2005. From July 2nd to July 10th a campaign involving three telescopes of the observatory, the WHT, the TNG and the NOT was driven by a group led by ING astronomer Javier Licandro. LIRIS at the WHT was used from July 3rd to 7th to obtain near infrared images in the J and K bands and near infrared spectra. Also another group lead by Stephen Lowry (QUB) used the INT and the Liverpool Telescope to follow the activity of the comet from July 1st to 7th. The five largest telescopes of the Roque de Los Muchachos Observatory were used simultaneously to track an astronomical experiment in an unprecedented way.

On July 3rd, 2005 the Deep Impact impactor probe successfully separated from its mother craft onto a trajectory that would plunge the probe into the nucleus of comet 9P/Tempel1 at a velocity of 10 kms^–1. At 05:44:36 UT on July 4th the impactor collided with the comet producing an impact of 19 GJ of kinetic energy and excavating a crater shaped by gravity.

The first aim of the campaign was to study the dust ejected by the impact by using the high S/N images obtained in the visible and near infrared, and the spectra in the near infrared where there are many features due to gas emission. The evolution of the intensity and colour of the dust gives important information on the size of the ejected grains, like their size distribution and ejection velocities. The second aim of the campaign was to measure possible variations of the gas emission by means of visible spectroscopy, to detect any possible new activity in case the impactor penetrated deep enough to meet the fresh ices below the dust mantle. This would evaporate part of them and expose ices to the sun-light generating a new active area.

The impact produced an ejecta cloud of dust easily seen in the images. The wonderful atmospheric conditions (all nights were photometric, and the seeing was as good as 0.4arcsec) allowed the astronomers to obtain a set of excellent images in particular in the R and J bands. At these wavelengths the images were sampling the reflected sunlight by the dust in the coma. The spectra also showed that the gas contribution was very low in particular in the near infrared. Some conclusions about the dust cloud ejecta follow:

• The dust ejected by the impact formed a semi-circular expanding cloud that extended from position angles (PA) 145° to 325°.
• Assuming an albedo typical of cometary grain size, the flux of the dust ejecta allowed us to estimate that the total mass of dust ejected was ∼10⁶ kg (equivalent to about 10 hours of normal comet activity).
• The orientation of the ejecta proves that the impact happened below the orbital plane of the comet.
• The position of the leading edge of the dust cloud present on the July 4th images show that it expanded outward at a projected speed of about 200±20ms^–1 (though varying with azimuth).
• In the following days the shape of the cloud changed because of the effect of solar radiation pressure that moved the dust particles to the tail of the comet (PA=110°). The maximum projected distance in the sunward direction, achieved on July 7th, was 30,000 km. By July 9th most of the ejected dust was moved to the coma and the comet looked like as in the pre-impact phase. The ejected dust is diluted in the comet tail.

The study of the structures of the dust coma in high S/N images provided also very interesting results:

• The comet presented some dust structures in the pre-impact phase that indicate that the nucleus had some particularly active regions.
• These structures remained after the impact, thus these active regions were not affected.
• The new structures observed after the impact on July 4th rapidly disappeared and none remained at a high S/N level after a few days.

Figure 1. First row: Sequence of calibrated near-infrared J-images of the dust coma of comet Tempel1 obtained with LIRIS at the WHT. Notice the changes in the images from July 3rd (pre-impact) to July 4th (taken 16 hours after the impact). Second row: Post-impact J-images processed to show only the dust ejected by the impact. Each image has been divided by the pre-impact one obtained on July 3rd. The evolution of the ejecta cloud is clearly seen. Third row: Same as 2nd row shown in a different flux scale. [ JPEG | TIFF ]

The observations from the INT were very important for completing the time base coverage of the comet as it fell below the sky from the primary observing site at Mauna Kea, Hawaii. The observing slot ran from July 1st to July 7th, 2005. A period which overlapped the Deep Impact encounter allowed the observers three nights pre-impact and four nights post impact observing. The strategy was to use the Wide Field Camera to obtain image mosaics up to 5 million kilometres along the projected anti-solar direction to look for ion-tail features that may have been produced as a result of the impact. The post impact observations quickly revealed that no such ion features were present, which was subsequently confirmed by other observers performing similar programs. With this in mind it was decided to focus on deep optical imaging of the central gas and dust coma through UBVr'i'O+ filters. When the comet was imaged on July 4th, about 16 hours after the impact, the comet was seen to have increased in brightness by a factor of two —as measured in the central pixel— compared to the July 3rd pre-impact levels. Some dramatic changes were seen in the dust coma. The Deep Impact event did not create a new period of sustained cometary activity, and in many ways the artificial impact resembled a natural outburst.

In conclusion, the impact was an impulsive event that affected the dust mantle of the comet. A large amount of dust was ejected into the coma in a very short time. In no more than 5 days this dust dissipated. Also, if the impactor reached the fresh-ices below the dust mantle, it did not excavate enough to expose a sufficient amount of ices to create a new region sufficiently active to be easily detected.

Figure 2. When comet Tempel1 came into view from La Palma, some 16 hours after the Deep Impact probe struck the comet, observes were able to start tracking the target comet with the INT. Both images above are a combination of 7×20 second Sloan-Gunn r' (red) filter images which isolate the dust component of the coma. The image on the left was taken on July 3rd between 21:56 and 23:03 Universal Time, about 7 hours before impact. The image on the right was taken between 22:08 and 23:56 UT on July 4th, 16 hours after probe impact. The comet was seen to increase in brightness by a factor of two —as measured in the central pixel— before and after the impact as seen from this location. Even in these images the effects of the impact can be seen by the changing coma shape between the two images. North is up and East to the left. The field of view in both images is 340×340 arcseconds, which is equivalent to ∼220,000×220,000 km at the comet. [ JPEG | TIFF ]

Figure 3. Larson-Sekanina image processing techniques were applied to the coadded r'-filter coma images from four of the nights to reveal the dramatic changes in the structure of the dust coma that resulted from the impact. Before impact on July 3rd (upper left panel), three jet structures can be seen to be emanating from the comet’s nucleus. On July 4th, just 16 hours after probe impact, the intensity of the jet/coma structures in the West and Southern direction increase dramatically (upper right panel). North is up and East to the left. On July 5th, new curved structures can be seen in the North-Western quadrant (lower left panel), possibly made visible by the extra material ejected into the coma from the impact. The structures seem to return to their pre-impact status on July 6th (lower right panel). The first three panels are directly comparable in terms of atmospheric seeing (variable seeing can complicate matters when comparing such processed images). The seeing on July 6th is actually much better than the other images. The images have been scaled to account for the slight extinction and geometry changes from night to night. Also, the images were taken under photometric conditions. Of course, nucleus rotation will need to be factored into this analysis, but even so, impact effects are still clearly seen in these images. The field of view in every image is 95×95 arcseconds, which is equivalent to ∼62,000×62,000km at the comet. [ JPEG | TIFF ]

Figure 4. These dramatic images of the expanding and dissipating ejecta plume were obtained by dividing the July 4th, 5th, and 6th coadded images of the comet by the pre-impact image on July 3rd. North is up and East to the left. The plume expands mostly into the South-Western quadrant, and appears to be decelerating at a non-uniform rate. The dust particles at the leading edge of the plume, are expanding at a rate of ∼210 ms–1 (±10%) on July 4th (measured at a Position Angle of 225°). The field of view in every image is 190×190 arcseconds, which is equivalent to ∼123,000×123,000 km at the comet.[ JPEG | TIFF ]

References:

• The Tempel 1 Observing Collaborators Team, 2005, "Deep Impact: Observations from a Worldwide Earth-based Campaign", Science, 310, 265.
• J. Licandro, 2005, "The Deep Impact Event as Seen from the Roque de Los Muchachos Observatory", ING Newsl., 10, 4.
• S. C. Lowry, A. J. Coates, A. Fitzsimmons, G. H. Jones, C. M. Lisse, 2005, "Deep Impact Observing at the Isaac Newton Telescope", ING Newsl., 10, 6.