Direct Hit on a Near Miss

John Davies

UKIRT Support Scientist, Joint Astronomy Centre, Hilo

In August, UKIRT demonstrated its flexibility and reliabilty by snatching high quality near-infrared spectra of a rapidly moving object visible briefly during the pre-dawn twilight.

At about lunchtime on August 10th Hawaii time (22:30 UT) the Earth had a close encounter as the recently discovered asteroid 1998 ML14 flew within 2.7 million kilometers of our planet. The asteroid, a member of the class of "Near Earth Objects" of which several hundred are known, had been discovered about 6 weeks earlier. This was too little warning for a regular PATT application but, by co-incidence, UKIRT was already scheduled to be observing asteroids that night as part of Andrew Rivkin's (Lunar and Planetary Laboratory, Arizona)  CGS4 study of hydrated minerals in main belt asteroids. As I was at the telescope with him anyway,  we decided to try to get some near-IR spectra of 1998 ML14. However the asteroid was a long way South (Dec -35) and so was not well placed for UKIRT, worse still, it would not rise into UKIRT's pointing envelope until  minutes before dawn. With the inevitable uncertainty in the orbit of a newly discovered object, especially one whose path was under the influence of the Earth's own gravitational pull, it promised to be a challenging observation.

Rather than risk losing time from our primary project preparing for an observation that might not even be possible, we planned to use the first night as a dry run to confirm that we could at least find the rapidly moving target. Just after the asteroid rose, and already nine minutes into nautical twilight, Chris Davis slewed UKIRT to the predicted position and within a few moments we had something in sight. Moments later the object's motion of  more than half an arc second per second in both RA and Dec confirmed the identification. Chris centred the asteroid in the guide box and turned on the autoguider. The fast-guider locked on and, even though it was now getting quite light, we started peaking up CGS4. This done we began integrating and to our great satisfaction managed to get a few minutes of data before the asteroid vanished in the  rapidly brightening sky.

Flushed with this sucess we decided to repeat the attempt the following night, but this time we set up our standard stars and pointing before the asteroid rose and ambushed it, locking on the guider within a minute of it coming over UKIRT's horizon and squeezing in spectra in two separate wavebands, plus a final standard, before dawn. Even though the asteroid was fading at a rate of about one magnitude every day (due to increasing distance and increasing solar phase angle as seen from Earth) we repeated the exercise the following morning and finished up with enough data to produce a high signal to noise spectrum from about 1.1 to 2.5 microns (Fig 1.)
 
 

FIGURE 1 : A preliminary composite spectrum of 1998 ML14. All five spectral fragments have been normalised and combined. The structure just longward of 2 microns is not reliable, it is due to poor atmospheric cancellation in this region.

Apart from the intrinsic value of this spectrum, which will enable us to identify specific minerals on the asteroid and help to classify it, these observations show how well UKIRT can operate as an observatory system. On the final night I kept notes of how the observing compared to our plan and the table below shows the result.
 
 

Activity	Planned Time (UT)	Actual Time (UT)
Begin FLAT-ARC for 1st standard	14:15	14:15
Slew to BS1532 & peakup	14:20	14:19
Start integrating on BS1532	14:25	14:26
Begin slew to asteroid & peakup	14:35	14:36
Start integration on asteroid	14:40	14:41
Slew to BS1532 & start integrating	14:55	14:57
Change grating to new wavelength	15:00	?
Slew back to asteroid, peakup & start integration	15:05	15:07
Slew to BS174 & start integrating	15:25	15:21
Do FLAT & ARC for 2nd grating setting	15:30	15:27

All of this was done at the end of a busy night, at an airmass of between 2.05 and 1.7,  in astronomical or nautical twilight, on a moving object of uncertain position and with full tip-tilt correction on the asteroid. To me this seems a pretty impressive technical performance and really shows just what UKIRT, plus its instruments and operations team, can achieve.

What did we learn? From the absorption band around 2 microns we deduce the prescence of the mineral pyroxene, which means 1998 ML14 is probably a S type, or stony, asteroid. From this we can guess the albedo (reflectivity) and estimate its diameter to be about 1km. Objects this size impact the Earth on average every 150,000 years and strike with an energy of about 100,000 megatons of TNT, dwarfing the largest nuclear weapon ever built. Had 1998 ML14 struck, it would indeed have been a Deep Impact.

And on the same subject...

Although from near-infrared spectra you can determine the spectral type of an asteroid, and hence get an idea of its albedo and so its size, if you want to determine the diameters of asteroids accurately, UKIRT can do much better than this. With only the reflected light to go on you can't really tell if you are looking at a large dark object (like a bit of coal) or a small white object (like a snowball). The way around this is to observe  in both visible light and at wavelengths of 10 and 20 microns. This works because some of the sunlight arriving at the asteroid is reflected, and you measure that in the visible, and the rest is absorbed and warms up the asteroid. This is then re-emitted as thermal infrared radiation which can be measured using instruments such as CGS3 and Michelle. Since the asteroid is in equilibrium with the sunlight it receives, from the ratio of the reflected (visible) and emitted (infrared) light you can determine both the size and albedo of the object.
 
 

FIGURE 2 : An example of thermal model fits to UKIRT data is shown, which is taken from Harris et al. (1998). The curves correspond to three different thermal models, after adjusting parameters such as diameter and albedo to give the best fit in each case. Of all the models the dotted curve, which corresponds to the near-Earth asteroid model mentioned in the text, fits the data best.

Alan Harris from the German Aerospace Centre's Institute of Planetary Exploration in Berlin is the PI of a UKIRT programme to do just this. Using CGS3 as a spectrophotometer and synchronising observations with optical observers around the world, he and I have sucessfully observed over half a dozen near-Earth asteroids since 1997. Alan has applied his new thermal model, which is designed specifically for observations of small objects at the high solar phase angles typical of those encountered when observing objects close to the Earth, to determine sizes and albedos of the asteroids (see Harris, A.W., Davies, J.K, Green, S.F. (1998). Thermal Infrared Spectrophotometry of the Near Earth Asteroids 2100 Ra-Shalom, and 1991 EE, Icarus, in press). We were unable to apply this technique to 1998 ML14 as CGS3 was not planned to be on UKIRT in August and it did not seem worth installing it for just a single observation. However, when Michelle is mounted at UKIRT it will be available for a good fraction of the time and will hopefully be used in its imaging or spectrophotometric modes for similar targets of opportunity. In fact, with Michelle and UFTI/UIST on at the same time, UKIRT will be able to do both reflected (JHK) and thermally emitted (NQ) radiation almost simultaneously, making it a powerful tool for performing these types of observations.

P.S : For those of you who have seen the film Armageddon don't worry, there is no known asteroid the size of Texas. For the Deep Impact fans you can worry a bit, the scenario enacted by this movie is fairly realistic, although unlikely in our lifetimes.
 

Back to Contents