Icarus 151, 139–149 (2001) doi:10.1006/icar.2001.6613, available online at http://www.idealibrary.com on Spectral Properties of Near-Earth Objects: Palomar and IRTF Results for 48 Objects Including Spacecraft Targets (9969) Braille and (10302) 1989 ML Richard P. Binzel Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 E-mail: [email protected] Alan W. Harris Jet Propulsion Laboratory, Pasadena, California 91109 Schelte J. Bus Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 and Thomas H. Burbine Department of Mineral Sciences, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560 Received July 24, 2000; revised February 1, 2001 INTRODUCTION We present results of visible wavelength spectroscopic measure- ments for 48 near-Earth objects (NEOs) obtained with the 5-m tele- One of the most fundamental goals in the study of primi- scope at Palomar Mountain Observatory during 1998, 1999, and tive bodies in the Solar System is to achieve an understanding early 2000. The compositional interpretations for 15 of these ob- of the relationships between asteroids, comets, and meteorites. jects have been enhanced by the addition of near-infrared spectra Because asteroids, comets, and the immediate source bodies for obtained with the NASA Infrared Telescope Facility. One-third of meteorites exist within the near-Earth object (NEO) population, our sampled objects fall in the Sq and Q classes and resemble ordi- a compositional investigation of this population provides the best nary chondrite meteorites. Overall our sample shows a clear tran- basis for understanding these relationships. In addition, the NEO sition between S-type and Q-type compositional classes over visible population poses a long-term collisional hazard to the Earth as and near-infrared wavelengths. Taken together these results point well as the future opportunity for utilization of space resources, toward an abundance of near-Earth asteroids capable of providing sources for ordinary chondrite meteorites. Our sampling strategy thus providing a highly important and practical rationale for its favors targeting the smallest observable objects and we report re- study. sults for the 15-m diameter object 1998 BT13, the smallest spectro- We report the results of our ground-based reconnaissance pro- scopically measured NEO to date. NEOs show a greater spectral gram of NEOs conducted using the 5-m (200-inch) telescope diversity than main-belt asteroids, and our small sample includes at Palomar Mountain Observatory from January 1998 through objects falling in the rare categories of K, L, O, and V classes. The K- January 2000. Our targets were chosen primarily from the sam- class object 1999 JD6 is found to match CV chondrite meteorites. ple of known NEOs that lacked previous spectroscopic mea- Potential spacecraft targets received top priority for observation, surements and from discoveries made just prior to or during with the ordinary chondrite-like composition of (9969) Braille be- our observing runs. Priority was given to those targets having ing reported prior to the Deep Space-1 encounter. The relatively the greatest H magnitude values, as these objects are presumed accessible asteroid (10302) 1989 ML displays a neutral spectrum to be smallest in size. We reached a limit of H = 26 for 1998 that may be interpreted as a shock-darkened ordinary chondrite. BT13, estimated to have a size in the range of 15–20 m (for c 2001 Academic Press Key Words: asteroids, composition; meteorites; surfaces, aster- albedo 0.25 to 0.15), smaller than the size of the Palomar 200- oids. inch dome where we were observing! (See Fig. 1.) We addition- ally performed directed measurements of objects identified as 139 0019-1035/01 $35.00 Copyright c 2001 by Academic Press All rights of reproduction in any form reserved. 140 BINZEL ET AL. FIG. 1. Schematic diagram for the smallest near-Earth object spectroscopically measured to date, 1998 BT13, which has an estimated size of only 15–20 m. At this size, 1998 BT13 is smaller than the dome of the 200-inch telescope where observations were made. destinations for spacecraft encounters: (9969) Braille as the tar- 0.92 m, thus providing roughly 0.09 m of spectral overlap get for Deep Space 1 (DS-1) and (10302) 1989 ML as a formerly with the blue channel. This overlap is possible since the dichroic designated target for MUSES-C. Although not on the target list does not reach maximum transmission before 0.62 m. Still, for any currently approved mission, interest in (10302) 1989 ML the reflectivity of the dichroic averages less than 10% over the remains high because of its relatively easy accessibility from interval of 0.52 to 0.62 m, so that the usefulness of this overlap Earth. region in matching the vertical scaling of the spectral halves is generally limited to the brighter asteroids. PALOMAR INSTRUMENTATION, OBSERVING, A relatively wide 6-arcsec slit was utilized to maximize the AND REDUCTION PROCEDURES amount of light being sent to the detectors from our faint pro- gram objects. The length of the slit was sufficient to provide a Our Palomar measurements consist of visible wavelength region of sky 50 arcsec wide on either side of the object spec- spectra obtained using the Double Spectrograph (Oke and Gunn trum. This allows for a proper subtraction of the sky during 1982). This instrument uses a pair of charge–coupled device the data reduction. The slit remained fixed in a N–S orientation (CCD) cameras to simultaneously record the blue and red halves throughout all observations, and great care was taken to make of a long-slit spectrum. To increase the signal-to-noise per pixel, all observations as close to the meridian as possible, minimizing and greatly reduce the readout time for our observations, the the effects of differential atmospheric refraction. The capability readout for each 1024 1024 CCD array was binned by a factor for the telescope to track at nonsidereal rates in both coordi- of 2 2, and only that fraction of each CCD image containing nates allowed for accurately maintaining fast-moving targets in the spectrum and surrounding sky was stored. Due to different the slit. This tracking was monitored and manually corrected focal lengths for the two CCD cameras, the resulting spatial as necessary through the use of a sensitive slit-viewing cam- scales for the blue and red channels are 1.25 and 0.94 arcsec per era. Under good seeing conditions, the limiting magnitude for binned pixel, respectively. measurements extends below V = 19. For our observations, separation between the blue and red Standard data-reduction procedures were followed using pack- channels was accomplished using the “D-52” dichroic, which ages contained in image reduction and analysis facility (IRAF). has a steep cutoff in transmission shortward of 0.52 m. For Preprocessing of the CCD images involved bias and flat-fielding the blue channel, we utilized the 300 line/mm grating (blazed at corrections, using calibration images taken nightly at the tele- 0.399 m) that gives a dispersion of 7 A˚ per binned pixel over scope. The procedures used to produce the extracted one-dimen- the usable spectral range of 0.36 to 0.62 m. For the red channel sional spectra are detailed by Bus (1999) and are essentially we utilized the 158 line/mm grating (blazed at 0.75 m) to give the same as described by Xu et al. (1995). Upon extraction, 10 A˚ per binned pixel dispersion over the range from 0.53 to each spectrum was rebinned to a uniform dispersion of 25 A˚ PALOMAR AND IRTF SPECTRAL MEASUREMENTS OF NEOS 141 per pixel. The normalization of these spectra to units of relative Light first passes through a 1.2-arcsec slit before entering the flat reflectance was provided by observations of bright solar analog face of the grism. Dispersed light from the grism then passes stars. For this program, we utilized the widely accepted stan- through a PK50 filter to block the transmission of thermal ra- dards Hyades 64 and 16 Cyg B as well as the United Kingdom in- diation to the thermally sensitive detector. The grism and the frared telescope (UKIRT) standard FS13 and Landolt 102-1081, PK50 reside within the CVF wheel of the NSFCAM system. fainter stars which we found to be comparable in their quality The wavelength range for usable spectra is determined by a pair as solar analogs. All objects were reduced in exactly the same of blocking filters residing within the NSFCAM filter wheel. manner and as a check, bright main-belt asteroids having pre- This combination consists of a short pass filter that transmits viously known spectral characteristics were also occasionally radiation shortward of 2 m and a long pass filter that trans- observed. All main-belt results proved consistent with previous mits radiation longward of 0.85 m. (All cut-on and cut-off measurements. These objects are included within Results. wavelengths are the half-power points for steep regions of the To accurately merge the blue and red halves of the calibrated spectral transmission curves.) The resulting light that reaches spectrum, a third-order polynomial was fit repeatedly over the the NSFCAM 256 256 InSb array covers the range of 0.85 wavelength range of 0.44 to 0.66 m and the vertical scaling to 2 m. Only the light from the first-order spectrum, covering of the blue half was adjusted with respect to the red half. The approximately 0.85 to 1.65 m is usable as the second-order best scaling factor was found when the value of chi-squared was spectrum begins contaminating the region 1.7 to 2.0 m.