Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.naic.edu/~astro/aotms/performance/PositionSwitching.pdf Дата изменения: Tue Dec 24 17:05:30 2013 Дата индексирования: Sat Mar 1 00:09:35 2014 Кодировка: Поисковые слова: п п п п п п c 1999 s4 linear

Position Switching Separation Tolerance with Arecibo in L-Band
Travis McIntyre
The University of New Mexico July 2010 Revised December 2013 Abstract The blocked telescope aperture at Arecibo creates standing waves in the bandpass around 1MHz. The size and shape of the standing waves depends on the altitude and azimuth of the telescope, and so the effectiveness of removing them through position switching depends on the angular separation in alt-az coordinates of the on- and off-source positions. Test observations with Arecibo in L-Band show that position switching effectively removes standing waves for separations of one half FWHM or less. Standing waves remain for separations of one FWHM, increasing the rms of the baseline by 20%. Introduction A telescope with a partially blocked aperture will reflect some incoming radiation through the telescope support structure, reaching the receiver at a later time and thus out of phase with the original wavefront. When the data is sampled by the spectrometer, its power spectrum will include a term that corresponds with the time delay of the reflected radiation, and so the Fourier Transform produces a standing wave in the bandpass. These reflections, and thus the strength and shape of the standing waves, are highly dependent on the altitude and azimuth angle of the telescope, particularly at Arecibo where most of the receiver structure does not move with the telescope. On short time scales, these standing waves are relatively static and can be removed with position switching (i.e. integrating the sky both on-source and off-source and subtracting the two). However, because the support structure at Arecibo is not polarly symmetric, the effectiveness of position switching depends on the separation angle, , in altitude and azimuth coordinates of the on- and off-source positions. Method In an effort to define an upper limit for , several observations were taken during June 2010 to test the effect of on-off positional difference on the quality of spectral baselines. A total of 12 blind position switched pairs were observed, 5 with = 0, 3 with = 1.7 (1/2 FWHM), and 4 with = 3.4 (FHWM). Observations were taken with the ALFA receiver using the Mock spectrometer with a bandwidth of 172 MHz centered on 1450 MHz. The band was divided into 8192 channels giving a spectral resolution of 21 kHz. Integration times ranged from 180 ­ 270 seconds. A position on the sky was integrated to create an `on-source' observation and then the same alt-az path was tracked across except offset by a separation in declination, . The observations were bandpass calibrated into flux (mJy) and the heliocentric velocity (km/s) for 21cm neutral hydrogen emission. A description of the resulting spectra can be seen in Figure 1. Each baseline was fit to first order and the noise was measured to quantify the quality of the baseline. Not all of the position switched pairs had the same total integration time, so the noise between observations was compared as a percentage of measured rms over expected rms. Expected rms is calculated from the radiometer equation, using a system temperature of 28 K, a channel width of 21 kHz, and the time of the integration in seconds. Measured rms is computed from each spectrums baseline with RFI and sources masked.


Figure 1. Example of the spectra taken for testing position switched separation tolerance. The spectrum shows flux in mJy versus heliocentric velocity of redshfited 21cm emission in km/s. Spectrum notations are labeled.

Results The results are shown in Figure 2. Measured rms averages 99.3% of expected rms for 0 and 1.7 (1/2 FWHM) separations, but worsens to 120% for 3.4 (FWHM) separations. An increase in the amplitude of standing waves for 3.4 separation is clearly seen. For a neutral hydrogen galaxy survey, it is not possible to fit and remove standing waves or clip them from the power spectrum because the width of the waves (100 - 200 km/s) is similar to the velocity width of galaxies (50 ­ 500 km/s). Therefore, an upper limit of 1.7 is chosen for for extragalactic neutral hydrogen observations in Lband.


Figure 2. Reduced spectra from ON-OFF separation test. Angular separation increases from left to right (0', 1.7')