1 Introduction

Determinations of the solar diameter with astrolabes have steadily been pursued by various groups (Laclare 1983; Leister & Benevides-Soares 1990; Sánchez et al. 1993; Noël 1995; Leister et al. 1996; Golbasi & Kilic 1997). Comparison of the absolute values from other instruments is difficult, due to unequal definitions of the observational solar limb, to systematic effects peculiar to each technique and, in many instances, to bias introduced by visual observations (Wittmann 1977; Gilliland 1981; Ribes et al. 1988; Débarbat & Laclare 1990; Laclare et al. 1996). The results presented here are a contribution towards keeping a record of the observed variations of the solar diameter, aimed to add to the astrolabes series and given the low latitude of the station at Rio de Janeiro. The observation of the Sun with astrolabe at the Observatorio Nacional/ CNPq (ON), Rio de Janeiro ($\phi = -22^\circ 53' 42.50^{\prime\prime}, \lambda = +2^{\rm h} 52^{\rm m} 53.479^{\rm s}$, h= 33 m) started in 1978 focused on the determination of Earth orbit parameters and on the orientation of the astrometric reference frame (Penna et al. 1996, 1997a). The observations were carried out visually with a 30$^\circ$transparent prism and enabled also measurements of the solar diameter. The standard deviation of the daily values for the 1978-1981 campaign was 0$\hbox{$.\!\!^{\prime\prime}$}$4 (Penna et al. 1997b), which is comparable with the outcome from other astrolabe stations under the same conditions. Moreover, the average density was greater than 100 observations/year, allowing, thus, to assert the site quality.

In this paper the results from the first seven months campaign, January to July, 1997, are presented. We analyze the results from the point of view of the dependence of several observational parameters, as the observation time of the day and transit length, zenith distance and heliographic latitude, in order to verify whether there are significant effects to be considered when the time series are regarded. The results are discussed in Sect. 3.

All the measurements were taken with a CCD camera for the imaging of the solar limb. The reductions are performed through a constant group of pre-programmed routines upon the digitized images. This ensures a fully "impersonal'' result. In the following section the instrument's present method of operation and principles of reduction are briefly described.

The observations touch the extremes of the Sun's declination and demonstrate the feasibility of the astrolabe observations of the solar diameter all year round, with high density of measurements.