Optimizing the reduction of radio-astronomical observations during adjustment of a multielement interferometer

1) Equations (8) imply that the variances of the correction estimates of the direction of the baseline B_i are independent of the configuration of the interferometer elements which do not constitute B_i, but they do depend on the number m+1. This fact can explain why the optimal measurement procedure requires the calculation of correlation functions in Eq (11) between all possible pairs of received signals [the number of these is m(m+1)/2], while 2m parameters must be evaluated (dimensionality of the vector Q). There are, consequently, superfluous baselines taking part in the measurement of the vector Q. Using the information derived from the "superfluous" baselines permits one to increase the accuracy of the measurements by roughly a factor of {Mathematical expression}. 2) Analysis of the cited results shows that the measurement of corrections to the baseline directions for an interferometer is possible down to an accuracy of a fraction of a lobe width. The process does not place stringent limits on the interferometer parameters. Lowering of T_en, broadening of the bandwidth 2F, and increasing the observation time T and number of interferometer elements are all entirely possible. Further improvement in the accuracy is limited by the randomly varying refractive index from point to point in the propagation medium, engendering amplitude and phase distortions in plane waves within the interferometer antenna apertures, and it is also limited by the accuracy with which the coordinates of reference sources are known. 3) Comparison of the calculated variance of estimated parameters in the example above shows that the accuracy of the estimates depends essentially on the moment at which the measurement is made. The right ascension of the baseline is most accurately measured at the moment when β-β₀ (the phase, equal to the difference in right ascension between the baseline vector and the cosmic radio source at t=0) is equal to zero.