Gravity-dependent signal path variation in a large VLBI telescope modelled with a combination of surveying methods

Abstract

The very long baseline interferometry (VLBI) antenna in Medicina (Italy) is a 32-m AZ-EL mount that was surveyed several times, adopting an indirect method, for the purpose of estimating the eccentricity vector between the co-located VLBI and Global Positioning System instruments. In order to fulfill this task, targets were located in different parts of the telescope’s structure. Triangulation and trilateration on the targets highlight a consistent amount of deformation that biases the estimate of the instrument’s reference point up to 1 cm, depending on the targets’ locations. Therefore, whenever the estimation of accurate local ties is needed, it is critical to take into consideration the action of gravity on the structure. Furthermore, deformations induced by gravity on VLBI telescopes may modify the length of the path travelled by the incoming radio signal to a non-negligible extent. As a consequence, differently from what it is usually assumed, the relative distance of the feed horn’s phase centre with respect to the elevation axis may vary, depending on the telescope’s pointing elevation. The Medicina telescope’s signal path variation ΔL increases by a magnitude of approximately 2 cm, as the pointing elevation changes from horizon to zenith; it is described by an elevation-dependent second-order polynomial function computed as, according to Clark and Thomsen (Techical report, 100696, NASA, Greenbelt, 1988), a linear combination of three terms: receiver displacement ΔR, primary reflector’s vertex displacement ΔV and focal length variations ΔF. ΔL was investigated with a combination of terrestrial triangulation and trilateration, laser scanning and a finite element model of the antenna. The antenna gain (or auto-focus curve) ΔG is routinely determined through astronomical observations. A surprisingly accurate reproduction of ΔG can be obtained with a combination of ΔV, ΔF and ΔR.