Astron. Astrophys. 344, L29-L32 (1999)

2. Observations and image processing

The monochromatic and simultaneous WL images (48 series) were acquired with the IPM (Cavallini 1998) at the THEMIS telescope (Teide, Tenerife) on August 20, 1997. The sequence consists of about 1h time series (between 7:21 and 8:24 U.T.) and refers to a quiet solar region at disk center. The telescope set-up limited the image quality to about 0.7 arcsec.

Two 512512 CCD cameras (Berrilli et al. 1997), binned to 256256, were used to record images with an exposure time of 200 ms. The image scale was 0.134 arcsec/pixel, for a total field of view (FoV) of about 3434 arcsec. This FoV was reduced to 2525 arcsec because of the telescope tracking loss.

The 48 series of monochromatic images refer to 7 and 9 spectral points, with a bandwidth of about 2.1 pm, respectively within the C I 538.03 nm and Fe I 557.61 nm photospheric lines and near continua. The C line images represent the lower photosphere ( 60 km) while the Fe line images represent the higher photosphere ( 370 km) above , as reported in Komm et al. (1991). The time separation between two subsequent images was about 2.5 s.

After the standard corrections for atmospheric transparence variations, dark current and flat field, we applied a "phase opposition" filter (Espagnet et al., 1993) to remove the 5-minute acoustic waves (p-modes) and separate granular and oscillatory fluctuations. The telescope tracking failure, occurred during the observations, does not allow us the use of a more sophisticated filter.

From any cube of corrected monochromatic images we produce a cube of line-profiles using a Lorentz-profile fit. To measure the radial velocities we derive, from pixel by pixel fitted lines, a center line intensity image and a corresponding velocity field. Selecting the frames acquired during the best seeing intervals we derive four center line images, and the corresponding velocity fields, both for C and Fe lines.

A Wiener filter is applied to the WL images, contemporaneous to the Fe and C intensity and velocity frames, in order to correct the degradation (i.e. blur) due to instrumental transfer function, and to reduce the noise (Fig. 1, left). In the parametric Wiener filter we use a gaussian, with = 0.4 arcsec, as point spread function and we set the signal-to-noise ratio to a value of 15.

Fig. 1. Granulation field (34 square arcsec). Left Panel: Wiener filtered WL negative image of 07:42:49 U.T. August 20, 1997. Right Panel: The skeleton defining the cell boundaries (lines), and the binary field defining the granules (spots) corresponding to the same WL image.

After all the corrections, we apply a finding process to WL image to calculate geometrical properties of granules and cells, the latter being defined as containing both a granule (or more) and its corresponding part of the intergranular lane. We extract (Fig. 1, right) the skeleton of dark intergranular lanes and assume this one as definition of cell boundaries (Berrilli et al., 1998). The granules (Fig. 1, right) are identified using a dynamical threshold (Florio & Berrilli, 1998).

© European Southern Observatory (ESO) 1999

Online publication: March 18, 1999
helpdesk.link@springer.de