Astron. Astrophys. 353, L37-L40 (2000)

3. Astrometry analysis

A goal of this research has been to calibrate the limitations of our standard astrometry procedure for 43 GHz, as well as to study the potential precision of the astrometric data at this frequency. Therefore, the data-reduction procedure for the 43 GHz observation deliberately followed the same steps as those used for the 5 GHz (G95) or 8.4 GHz observations (G98; R99). We briefly go again over each step of this analysis: For our 43 GHz data, (i) we predicted, via a precise theoretical model of the geometry of the array and the propagation medium, the number of cycles of phase between consecutive observations of the same source to permit us to "connect" the phase delay (e.g. Shapiro et al. 1979; G95; R99); (ii) we defined as reference points in the 43 GHz images of the two radio sources the maximum of the brightness distribution (components Q1 in the maps of QSO 1928+738 and BL 2007+777; see Fig. 1) and subtract the contribution of the structure of the radiosources, with respect to the reference points selected, from the phase delays; (iii) we formed the differenced phase delays by subtracting the residual (observed minus theoretical values) phase delay of BL 2007+777 from the previous observation of QSO 1928+738; and (iv) we estimated the relative position of the reference points from a weighted-least-squares analysis of the differenced residual phase delays. For this analysis we used an improved version of the program VLBI3 (Robertson 1975).

In step (i), the geometry of our theoretical model (set of antenna coordinates, Earth-orientation parameters, and source coordinates) was consistently taken from IERS (IERS 1998 Annual Report, 1999). The theoretical model also accounted for the effect of the propagation medium in the astrometric observables. We modeled the ionospheric delay by using total electron contect (TEC) data from GPS-based global ionospheric maps generated at the epoch of our observations by the Center for Orbit Determination in Europe (Schaer et al. 1998). We followed the geometric corrections described in Klobuchar (1975) and Ros et al. (2000). We modeled the tropospheric zenith delay at each station as a piecewise-linear function characterized by values specified at epochs one hour apart. We calculated a priori values at these nodes from local surface temperature, pressure, and humidity, based on the model of Saastamoinen (1973). The antenna elevations were always higher than 20^o at all stations; this allowed us to use the dry and wet Chao mapping functions (Chao 1974) to determine the tropospheric delay at non-zenith elevations for each observation at each site. We estimated the tropospheric zenith delay at the nodes of each station, along with the relative position of the sources, from a weighted-least-squares analysis.

© European Southern Observatory (ESO) 2000

Online publication: January 18, 2000
helpdesk@link.springer.de