2 Results and discussion

The period of observations can be divided in two, before and after January 1st, 1998, when the filter after the image vehicle was swapped by one with a narrower bandpass. The error of a single observation of the solar semi-diameter, in the first part, is 0$\hbox{$.\!\!^{\prime\prime}$}$63 for the East and 0$\hbox{$.\!\!^{\prime\prime}$}$64 for West observations. In the second part the errors became smaller, and are 0$\hbox{$.\!\!^{\prime\prime}$}$57 and 0$\hbox{$.\!\!^{\prime\prime}$}$61, for the East and West observations correspondingly. With the new filter the width of the measured borders of the Sun became narrower and, as such, the definition of the limb of the Sun became more precise and the error of the diameter measurement decreased a little. Figure 1 brings the histograms of the observed semi-diameter values distribution in these two half-periods.

  

Figure 1: Histogram of the observed semi-diameter distribution. Solid line refers to the observations made before January 1st, 1998 while the dashed line refers to those made later. The plot demonstrates clearly that the errors of observations became smaller after the narrower bandpass filter was installed

There is a difference between the observations quality from East and West transits. The average temperatures for East and West observations depend on the season and vary from +25$^\circ$C in winter to +45$^\circ$C, and sometimes even higher, in summertime. However, the air temperature in the morning is some 6 to 8$^\circ$C lower than after midday, while the variation of the temperature in the morning (during East observations) is typically 4$^\circ$C higher than after midday (West observations). The agitation of the images is much smaller in the afternoons than in the mornings. Even the visual inspection of the images on the monitor during the observations reveals the effect.

Fried's parameter r₀ (Irbah et al. 1994) is, thus, larger for West observations (4.2 cm), than for the East ones (3.8 cm). That is, the seeing conditions are better after midday, the atmosphere being more stable at that time. The understanding of the process delayed the start of West observations till April 1997.

Normally, about 20 or more, up to 30, observations, on each side of the Sun's meridian cross, are made in summer months. These figures fall to about 10 or less observations, on each side, in June when the Sun is comparatively low above the horizon. The summer months, December to February, are mostly rainy at Rio de Janeiro, so during these months we have usually less observational days than in winter (June to August) but much more diameter measurements. In total there were 246 days of observation till June 30th, 1998.

The error for a single observation is slightly larger than the one obtained at the Calern Observatory (0$\hbox{$.\!\!^{\prime\prime}$}$3) owing to the environmental conditions of the ON site. The average daily error of the mean observed value of the solar semi-diameter for the first half-period is 0$\hbox{$.\!\!^{\prime\prime}$}$21, both for the eastern and western observations. In the second half-period the daily error became smaller and is 0$\hbox{$.\!\!^{\prime\prime}$}$18 for the East and 0$\hbox{$.\!\!^{\prime\prime}$}$19 for the West observations.

  

Figure 2: Values of the Sun semi-diameter vs. Modified Julian Day observed at the ON. East observation - open circles and West ones are black circles. Each point represents a mean daily value of the Sun semi-diameter with its average error i.e. average error bar of $\pm$0$\hbox{$.\!\!^{\prime\prime}$}$20

Figure 2 gives the values of the solar semi-diameter reduced to the one astronomical unit plotted against the Modified Julian Day. Each point represents the value of the mean daily semi-diameter. East and West series are differently represented. The larger scatter of the initial observations is evident, although no systematic offset is apparent. On the contrary, the average values for the 1997 and 1998 summer observations agree well.

  

Table 1: Monthly average semi-diameters obtained at ON (in arcsec) from January 1997 to June 1998. The "days" column displays the number of observational days during the month, while the "num.obs." column brings the total number of observations during the month

In the Table 1, number of the days of observation, total number of observations during a month and monthly averages are presented, both splitting East and West observations and combining them together. An analysis on the average values was already presented (Jilinski et al. 1998) and a forthcoming paper discussing the jointly results obtained in Calern and at ON is in preparation.

The complete results of the daily series can be retrieved in electronic form from the SIMBAD database and from the homepage of the Observatório Nacional (http://obsn.on.br/radius/). For the electronic form, the complete daily series are given: the number of observations, the semi-diameter mean value, its rms error, the date, the Modified Julian Date and the error for a single observation. The data are given for eastern and western observations separately.

In spite of the hard weather conditions of the observations, the instrument performed rather stably. No dependence was found for the observed semi-diameter on zenith distance, time length of observations (that vary from 2 to 7 min), azimuth or heliographic latitude of CCD scan across the Sun disc.