No abstract

The Mark III very-long-baseline interferometry (VLBI) system allows recording and later processing of up to 112 megabits per second from each radio telescope of an interferometer array. For astrometric and geodetic measurements, signals from two radio-frequency bands (2.2 to 2.3 and 8.2 to 8.6 gigahertz) are sampled and recorded simultaneously at all antenna sites. From these dual-band recordings the relative group delays of signals arriving at each pair of sites can be corrected for the contributions due to the ionosphere. For many radio sources for which the signals are sufficiently intense, these group delays can be determined with uncertainties under 50 picoseconds. Relative positions of widely separated antennas and celestial coordinates of radio sources have been determined from such measurements with 1 standard deviation uncertainties of about 5 centimeters and 3 milliseconds of arc, respectively. Sample results are given for the lengths of baselines between three antennas in the United States and three in Europe as well as for the arc lengths between the positions of six extragalactic radio sources. There is no significant evidence of change in any of these quantities. For mapping the brightness distribution of such compact radio sources, signals of a given polarization, or of pairs of orthogonal polarizations, can be recorded in up to 28 contiguous bands each nearly 2 megahertz wide. The ability to record large bandwidths and to link together many large radio telescopes allows detection and study of compact sources with flux densities under 1 millijansky.

No abstract

Aims. The Atacama Large Millimeter Array (ALMA) is an international astronomy facility to be used for detecting and imaging all types of astronomical sources at millimeter and submillimeter wavelengths at a 5000-m elevation site in the Atacama Desert of Chile. Our main aims are: describe the correlator sub-system which is that part of the ALMA system that combines the signal from up to 64 remote individual radio antennas and forms them into a single instrument; emphasize the high spectral resolution and the configuration flexibility available with the ALMA correlator. Methods. The main digital signal processing features and a block diagram of the correlator being constructed for the ALMA radio astronomy observatory are presented. Tables of observing modes and spectral resolutions offered by the correlator system are given together with some examples of multi-resolution spectral modes. Results. The correlator is delivered by quadrants and the first quadrant is being tested while most of the other printed circuit cards required by the system have been produced. In its final version the ALMA correlator will process the outputs of up to 64 antennas using an instantaneous bandwidth of 8 GHz in each of two polarizations per antenna. In the frequency division mode, unrivalled spectral flexibility together with very high resolution (3.8 kHz) and up to 8192 spectral points are achieved. In the time division mode high time resolution is available with minimum data dump rates of 16 ms for all cross-products.Key words. techniques: interferometric -techniques: spectroscopic -instrumentation: interferometersinstrumentation: spectrographs IntroductionThe Atacama Large Millimeter Array (ALMA) is an international astronomy facility. ALMA is a partnership between Europe, North America and Japan, in cooperation with the Republic of Chile (see Alma Information and note at the end of this paper). ALMA will consist of a main array of up to 64 12-m diameter antennas supplied by Europe (ESO) and North America (NRAO/AUI) and a compact array of twelve 7-m plus four 12-m antennas supplied by Japan (NAOJ). The main goal of this paper is to describe the essential features of the ALMA correlator being constructed for the main array (64 antennas) and to emphasize its huge versatility making this correlator a key element of all future astronomy programs led with ALMA. The ALMA correlator is a very large digital system that combines the outputs of up to 64 array elements using an instantaneous bandwidth of 8 GHz in each of two polarizations per antenna, and produces a single astronomical telescope from the 64 movable antennas distributed within a diameter of about 150 m to an expanded configuration of maximum antenna separation 18.5 km.A digital correlator system is the heart of any modern radio astronomy system in which the signal power is detected by measuring (a) the cross-correlation of all antenna pairs in an array of antennas, and/or (b) the auto-correlation of one (or all) single antenna(s). Digital lags can easily be implemented...

Analysis of 211 very long baseline interferometry observing sessions carried out between November 1979 and August 1984 has yielded estimates of the distances between various radio telescopes located inNorth America and Europe. The average rate of change of the distances between four radio telescopes in North America (Haystack Observatory, Massachusetts' Westford Radio Telescope, Massachusetts; George R. Agassiz Station, Texas' and Owens Valley Radio Observatory, California) and one in Europe (Onsala Space Observatory, Sweden) obtained from the analysis of these data is 19 + 10 mm/yr, where the (68% confidence interval) standard deviation is for the estimate of the rate of change of the Haystack-Onsala baseline length, the one determined most accurately from these data. This estimate of the standard deviation is dominated by the effects of correlated systematic errors due mostly to enors in the model used for the atmospheric delay which we infer introduces errors in each baseline length estimate of 40 mm standard deviation and 60 days correlation time. (By contrast the statistical standard deviation is only 2 mm/yr.) The estimated geologic rates of change of these baseline lengths, averaged over • 10 6 years, are 15 to 17 + 3 mm/yr for the various North American sites to Onsala. Improvements in our model of the atmosphere, and continued monitoring of the distances between North American and European telescopes, will allow the uncertainty of the rate estimates to be reduced over the next few years to a value small compared to our estimated rate of change of these baseline lengths. The use of multiple radio telescopes in Europe will allow us also to separate possible local site motions from plate motions. Paper number 5B5628. 0148-0227/86/005 B-5628 $05.00 and (2) to present the results for the baseline length estimates, their rates of change, and their uncertainties. The effects of the earth's atmosphere on the accuracy of these results are considered in detail by Davis et al. [1985]. 2. MATHEMATICAL AND STATISTICAL MODELS S-439 00 Onsala, Sweden.