We have searched for effects of quality decay of images due to the use of the the active mirror for the telescope’s fine pointing, as well as for the compensation of images motion due to atmospheric seeing. To this aim, we have examined the outline of the average solar edge, obtained by the images of the selected sample, on the three wavelengths of observation.
In Figure 9 we show the derivative of intensity along the solar edge, roughly estimated through the finite differences for intervals dr/R=0.01, in the three wavelengths of observation. The three curves thus obtained indicate the blurring of the net outline expected for the edge. These three curves are well described in terms of Gaussian curves with (2sigma) < 2 pixel (4 arcsec) for the images of the continuous red, and < 2.5 pixel (5 arcsec) for the images in the continuous blue and CaII K, 1024×1024. These results confirm the correct operation of the compensation system of the active mirror in the expected working conditions.

FIGURE 9: Outline of the derivative of the average intensity in the images in the continuous blue (+), red (x) and in the CaII K line (*) for intervals dr/R=0.01.

The optical design of the PSPT telescope, as well as the components employed, enable us to acquire images characterized by a nominal spatial scale equal to 1 arcsec/pixel. Checks made after installing the large-format CCD camera onto the telescope in 1997 had suggested that the acquired 2048×2048 images ( each one 1 arcsec/pixel, 16 Mb large), might be reduced to the 1024×1024 format ( each one 2 arcsec/pixel, 4Mb large) during the acquisition stage, without losing spatial information for the structures observed upon the solar disk. This choice, mainly due to the features of the telescope site, namely the hill of Monte Mario in Rome, was also affected by the limited archiving and computing resources available at the time. The format reduction was made straight into the stage of image reading, with the necessary initializing procedures of the CCD camera, thus making data archiving easier and saving a remarkable time in calculating instrument calibrations of acquired images.
In order to assess the information contents of images in terms of spatial resolution of acquired data, we applied Fourier’s bi-dimensional spectral analysis to sub-arrays extracted taken from the sample of analyzed images.
In Figure 10 we show an average outline of the bi-dimensional spectrum of power obtained from the analysis of the sample of images acquired in Roma, Monte Porzio and in Mauna Loa in the 1024×1024 format.


We can observe that the spatial information in the images are the same for all frequencies lower than the sampling frequency, or for the spatial scales larger than 1.9 arcsec, for a solar disk with a radius of 500 pixels. This suggests that the increase in the range of sampling from 1 arcsec/pixels for 2 arcsec/pixel, made by reducing the size of the acquired image during reading fase of the CCD, can lead to a loss of the information content of images.



FIGURE 10: average outline of the bi-dimensional spectrum of power obtained from the analysis of the sample of images acquired in Roma Monte Mario (dot), in Monte Porzio (continous line) and in Mauna Loa (dashed line) in the 1024×1024 format. From top to bottom respectively for image in CaIIK line, continous Blue and Red.