Calibration

The relative scarcity of sources whose absolute fluxes are known at these frequencies makes calibration of any microwave experiment a challenging proposition, as does the generic steepness of most radio source spectra. Additionally, of the sources which are well studied, few are accessible from the South Pole (and planets, when they are visible at all, never rise more than $23\fdg 5$ above the horizon). Moreover, with DASI's 20-cm apertures, beam dilution makes it inherently difficult to detect all but the brightest point sources. The gain of the telescope is approximately $10~{\rm\mu K/Jy}$, so that for a single baseline, to achieve a signal/noise $\sim$ 1 on a 1 Jy source requires about 3 hours of observation.

 

Figure: DASI images of calibration sources. Eta Carinae is shown at top left, with the central complex suppressed so that the substructure can be seen. The sharp fall-off of the flux with u-v radius (top right) degrades the accuracy of long baseline calibrations. The source PKS B0857-473 (bottom left) is much more compact, and is correspondingly brighter than Eta Carinae for baseline lengths $> 80\lambda$.

Accordingly, the absolute calibration of DASI was initially based on observations of Eta Carinae, a well-known Galactic HII complex. While point-like sources are more convenient, a source of arbitrary complexity can in principle be used as a flux calibrator, provided its flux can be determined for each baseline. DASI's simple horn design allows an accurate calculation of the theoretical aperture efficiency, and fluxes for the source are based on absolute load calibrations transferred to Eta Carinae in February 2000.

Eta Carinae's free-free spectrum makes it a good high frequency calibrator; the source is bright enough that its flux can be measured to a few percent on most baselines in approximately 15 minutes. As can be seen from Figure 3, however, the source is dominated by a bright central region, with much weaker flux on smaller scales, necessitating significantly longer integrations on the longest baselines to achieve uniform accuracy. These considerations have led us to transfer the Eta Carinae calibration to PKS B0857-473, a more compact Galactic HII region also observed by the BOOMERANG experiment (de Bernardis et al. 2000). Although its integrated flux is lower than that of Eta Carinae, as can be seen from Figure 3, the source is readily detectable, and uniformly bright even at our longest baselines.

Observations of the phase stability while tracking PKS B0857-47 demonstrate that it is sufficiently point-like (or at least radially symmetric), and that the background is sufficiently uniform, that the source is also suitable as a phase calibrator.

Recently, simultaneous DASI and CBI observations of a bright point source visible both from the South Pole and Cerro Chajnantor in Chile, have established a common flux scale between these experiments, effectively linking the DASI calibration to the brightness temperature of Jupiter.

Go back to DASI home page