Data Set Overview : All observations were obtained with SpeX, the low- to medium-resolution near-IR spectrograph and imager at the NASA Infrared Telescope Facility (IRTF) on Mauna Kea, Hawaii. This instrument uses prism cross-dispersers and gratings to achieve resolving powers up to R~2500 across the wavelength region from 0.8 to 5.5 microns. An infrared slit-viewer allows for object acquisition and guiding. The SpeX spectrograph employs a 1024x1024 Aladdin 3 InSb array and the slit-viewing imager uses a 512x512 Aladdin 2 InSb array. [RAYNERETAL2004]. Of specific importance to asteroid studies, SpeX provides a low-resolution (R~100 when using the nominal 0.8 arcsecond slit) 'prism' mode covering the wavelength interval 0.8-2.5 microns. In prism mode, asteroids as faint as Vmag~16.5 are routinely observed, with even fainter magnitude limits being reached when good sky conditions permit.  Asteroid observations are normally obtained by taking nodded spectral image pairs of the target asteroid, solar-analog stars, and (possibly) telluric standard stars, followed by calibration flat-field and arc-lamp images. The image pairs are taken by shifting the asteroid (or star) position on the slit (nodding the telescope position) by a fixed amount. In this 'A-B' image pair, the background sky spectrum is being recorded simultaneously with the target, allowing the sky to be subtracted in both the A-B and B-A pairs. After flat fielding the individual SpeX images and carrying out the pairwise sky subtraction, multiple images of the target or standard star can be stacked, and a 1-dimensional spectrum extracted, wavelength calibrated and re-binned as appropriate. Calibration of each asteroid spectrum in dimensionless units of relative reflectance is achieved by dividing the asteroid spectrum by that of a solar analog star.  Because this data set contains results derived by several different research groups, there are some variations in observing strategies and reduction procedures. This is particularly the case in the final stages of spectral calibration where effects of the Earth's atmosphere (telluric absorptions) are fitted and removed. Different techniques for telluric correction include: 1) the modeling of atmospheric absorption bands using ATRAN [LORD1992] which is based on the HITRAN database of molecular absorption lines (for convenience, a .pdf copy of Lord 1992, NASA Technical Memorandum 103957 is included in the document directory), 2) using the method included in SpeXTool [VACCAETAL2003] where observations of an A0 (telluric standard) star are convolved with a high-resolution model of Vega to construct a telluric correction spectrum, and 3) the use of SpecPR, where wavelength-dependent extinction coefficients are derived from nightly standard star observations. While the application of any of these telluric calibration methods leads to very similar results, users of this data set are encouraged to refer to the original papers for specific details about the data reduction and calibration procedures used.  Papers for which data are archived are: Binzel et al. (2001) M&PS 36, 1167-1172 [BINZELETAL2001] Clark et al. (2004) AJ 128, 3070-3081 [CLARKETAL2004] Rivkin et al. (2005) Icarus 175, 175-180 [RIVKINETAL2005] Sunshine et al. (2004) M&PS 39, 1343-1357 [SUNSHINEETAL2004] Sunshine et al. (2007) M&PS 42, 155-170 [SUNSHINEETAL2007B] Sunshine et al. (2008) Science 320, 514-517 [SUNSHINEETAL2008]  Data File Description : Data file names have the format: AAAA_YYMMDDTHHMMSS.tab, where AAAA is the permanent number or provisional designation of the asteroid, YYMMDD is the UT start date, and HHMMSS the UT start time of the observation. This naming convention is particularly useful when a study involved rotationally resolved spectra, where multiple spectral files are being archived for a single asteroid.  To determine what asteroid spectra have been archived, and to help navigate to specific files, a science index (science_index.tab) is included in the upper level of the data directory. This file lists all spectra in order of asteroid designation, and includes the data path, UT start and end times of the first and last asteroid observations used in creating the spectrum, the apparent magnitude, phase angle and heliocentric distance of the asteroid at the time(s) of the observations(s), and a listing of the solar analog stars used in calibration.  Each spectrum file contains 3 columns: wavelength (in microns), relative reflectance, and error in relative reflectance. For those spectra reduced using IRAF, the output spectrum is usually re-binned to a consistent dispersion, which allows different spectra to be easily compared and combined. During the various reduction and calibration steps, if a particular spectral data point is identified as discrepant, that point is flagged and removed from the final spectrum file. All spectra have been scaled by fitting a polynomial over the region centered on the J-band (1.215 microns), and then normalizing the fitted value at 1.215 microns to 1.00.  Thumbnail plots of all spectra archived in this data set are provided (pdf files) in the document directory. These plots are organized by publication (one file per publication), using the same naming scheme as the data directory structure.

Confidence Level Overview : The uncertainty for each spectral channel has been derived using Poisson statistics, and was based on the signal from both the object and sky, as well as the detector characteristics. The errors listed in each spectral file do not include systematic uncertainties. Examples of such systematics include uncertainties in the telluric absorption correction model, and uncertainties in the telescopic observations that can affect the average slope as measured over the length of the spectrum. By measuring several calibration stars over the course of a night and averaging the results from each asteroid / star combination, the level of systematic errors in the final asteroid spectra have been minimized.