The Balloon-borne Large Aperture Submillimeter Telescope (BLAST) 2006: Calibration and Flight Performance

The December 2006 flight of the Balloon-borne Large Aperture Submillimeter Telescope (BLAST) incorporated a 1.8-m parabolic primary mirror and large-format bolometer arrays operating at 250, 350, and 500 ñm. A complete description of the BLAST instrument can be found in Pascale et al. (2008). The BLAST bands sample the peak of the spectral energy distribution (SED) for cool dust (∼ 10 K). Astronomical signals at these wavelengths are difficult or impossible to access from even the best ground-based sites. As a result, BLAST has the ability to conduct unique Galactic and extragalactic submillimeter surveys with sub-arcminute resolution and high sensitivity. BLAST's primary scientific motivations are to study the spatial and redshift distribution and evolution of high-redshift star-forming galaxies and to probe the earliest stages of star formation within Galactic molecular clouds.

BLAST conducted a 250-hour flight, launching from McMurdo Station, Antarctica on 2006 December 21, and landing on the Antarctic Plateau 2007 January 2 (BLAST06). BLAST flew at an average altitude of 38.6 km with diurnal variations between 37.5 and 39.6 km.

Several extragalactic and Galactic fields were mapped, including two large (8.7 deg 2 ) and one deep, confusion-limited (0.8 deg 2 ) extragalactic fields and two large, overlapping regions (a 50 deg 2 deep region and a 200 deg 2 wide region) in the direction of Vela (Devlin et al. 2009;Netterfield et al. 2009).

The primary science goals of the BLAST experiment demand an absolute flux calibration accuracy of 5-10% in all three BLAST pass-bands. In particular, a target, uncorrelated 5% uncertainty is driven by the extra-galactic science case to enable precise measurements of colors and thereby constrain the bolometric luminosities and starformation rates of distant galaxies (Hughes et al. 2002). In order to achieve these goals, a primary calibration source for BLAST with the following properties was required: (i) point-like and bright enough to be detected in each band with a SNR exceeding 20σ in a single map; (ii) the absolute (correlated) uncertainty in the SED had to be less than 10%; and (iii) the uncorrelated components of the uncertainty in the SED (i.e. uncertainties in the ratios of flux densities in different BLAST bands) could be no greater than 5%. In this paper, we report on the calibration and performance of BLAST06, concentrating on the differences from the calibration procedures used in BLAST05, discussed in Truch et al. (2008) (hereafter T08). Section 2 outlines the basic reduction steps and characterization of BLAST06 data. Section 3 discusses the performance of the warm optics in BLAST06. Section 4 outlines the pointing performance of BLAST06. Section 5, the bulk of this paper, describes in detail the absolute calibration (from detector Volts to Jy on the sky) derived from the primary flux calibrator VY CMa (Hoffmeister 1931;Guthnick & Schneller 1939).

The data reduction for BLAST is discussed in detail in Pascale et al. (2008), Patanchon et al. (2008), and T08. Briefly, the data from BLAST consist of a set of 288 bolometer time streams, in voltage units, sampled at 100 Hz. Eighteen of these time streams are diagnostic channels, useful for removing common mode noise, the remaining 270 are coupled to the telescope. These bolometer data are first cleaned for post-flight analysis by being de-spiked and then deconvolved to remove the effects of the data acquisition system filters from the timestreams. The cleaned data are then combined with a post-flight pointing solution (Pascale et al. 2008) to make maps at each wavelength. The map-making process takes advantage of the multiple detectors, as well as significant scan cross-linking, to minimize striping due to instrumental drifts (Patanchon et al. 2008).

The bolometers in each array are corrected for relative gains, or flat-fielded, so that multi-bolometer maps can be generated. The flat-field corrections are determined by making individual maps for each bolometer from a single point-source calibrator, in this case, VY CMa (see 5.1). The bolometers are also corrected for responsivity variations over time by using the signals from a calibration lamp in the optics box which was pulsed every 15 minutes throughout the observations. The resulting signal is used to correct any time-varying changes in responsivity per bolometer. Both the time varying changes and the variations of beams across each array are small and the amount of variation is comparable to those detailed in 2 of T08 for BLAST05.

To calculate the flux density from a point source we adopt a matched-filtering technique similar to that used to extract point sources from several recent extragalactic submillimeter surveys (e.g., Coppin et al. 2006;Scott et al. 2006) and detailed in 2 of T08. The beam profile on the sky, or point spread function (PSF), used for calibration and flux extraction is generated by stacking and averaging several observations of VY

…(Full text truncated)…

This content is AI-processed based on ArXiv data.