Appendix D: CCD narrow band photometry  

Images of NGC 128 through the interference filters H$\alpha$ and the nearby continuum have been obtained at the 1.5 m Danish telescope during the same ESO run of the optical observations. The adopted filter has a width of 50 Ang and takes into account the redshift of the galaxy. The width is large enough to include the contribution of H$\alpha$ and [NII] lines.

After the procedures of bias, dark, flat-field correction, cosmic rays cleaning, sky subtraction, normalization at 1 s exposure, reduction to a standard airmass, and alignment of the images, we faced with the probelm of isolating the H$\alpha$ contribution with respect to the continuum. The problem is to find the constant factor $\alpha$ which equates the level of the continuum of the two images (H$\alpha$ and nearby continuum):

$$I_{\rm l}(x,y) = \alpha I_{\rm c}(x,y)$$ (D1)

where $I_{\rm l}$ and $I_{\rm c}$ give respectively the fluxes measured through the lines and the continuum.

In order to derive $\alpha$ it is convenient to establish the region of the frame where one can exclude the contribution of the emission lines. This is the region far from the center of the galaxy. The $\alpha$ coefficient could be obtained from the growth curves of the two images because the gradient of these curves scale with the same $\alpha$factor:

$$\partial_\rho \mu_{\rm l}(\rho) = \alpha \partial_\rho \mu_{\rm c}(\rho).$$ (10)

By increasing the radius of integration the contribution of the emission is rapidly vanishing. It is therefore sufficient to calculate $\alpha$ for large values of $\rho$ by excluding the central region.

In order to calibrate the H$\alpha$ flux we used the standard star L870-2 of the Oke (1974) catalog that was observed the same night. The flux F_* of this star at $\lambda$6654.29, corresponding to our recession velocity of 4180 km s^-1, is $1.933\ 10^{-14}\ \pm\ 0.024\ 10^{-14}$erg s^-1 cm^-2 Ang^-1. By adopting the previous procedure of analysis to the standard star, one obtains the ratio between the standard and measured flux $S = F_*/I_* = 1.14\ 10^{-15}\ \pm\ 0.05\ 10^{-15}$ erg cm^-2 ADU^-1, which express the sensitivity of our system. Then, by building the growth curve of the "pure'' H$\alpha$ image of the galaxy we get a total flux $F_{\rm g} = 3.08\ 10^{-14}\ \pm\ 0.61\ 10^{-14}$ erg s^-1 cm^-2, from which the resulting H$\alpha$ + [NII] luminosity turns out of $2.18\pm0.41\ 10^{40}$ erg s^-1, a value comparable to that of other S0 galaxies of the same luminosity (Buson et al. 1993).