4 Continuum and absorption line measurements

Figure 4: Illustration of the continuum determination procedure for the nuclear spectra of a) the non-emission line galaxy II Zw 82, and b) the star-forming galaxy Haro 1. The continuum pivot points are marked by the filled triangles. Vertical lines at the bottom of each panels indicate the wavelength windows used to measure the equivalent widths of Ca K and H, H$\delta$, CN, G band, Mg and Na lines (See in Table 5).

The equivalent widths of absorption lines and the continuum colors provide information about the stellar populations and physical parameters of galaxies. One of our goals is to study the star formation history and chemical evolution of BCGs. To this end, we also determined a pseudocontinuum at selected pivot points and measured the equivalent widths of 7 absorption features, integrating the flux within each window between the pseudocontinuum and the spectrum. The absorption features and continuum points are chosen based on the population synthesis method (Cid Fernandes et al. 2001) that will be used in a forthcoming paper.

4.1 Continuum measurements

In order to derive a continuum, we have measured the flux values at 9 pivot points, 3660, 4020, 4510, 4630, 5313, 5500, 6080, 6630, 7043 Å which were chosen to avoid regions of strong emission or absorption features (Bica 1988; Kong & Cheng 1999; Saraiva et al. 2001). The corresponding fluxes were determined as averages in 20 Å bins centered in the listed wavelengths. The determination of the continuum has been checked interactively, taking into account the flux level, the noise and minor wavelength calibration uncertainties as well as anomalies due to the presence of emission lines. The excellent quality of the spectra allowed a precise determination of the continuum in the majority of cases.

In addition, 3 point flux values (3784, 3814, 3918 Å) were measured, which were necessary for the determination of the continuum in galaxies with strong contribution of late B to F stars which present several high-order Balmer absorption lines in this region (Bica et al. 1994). Due to the crowding of the absorption lines, it is difficult to make automatic measurements. We thus selected the highest value of $\lambda$3784, $\lambda$3814 and $\lambda$3918 Å fluxes to represent the continuum in these spectral regions.

Figure 4 illustrates the application of the method to two of the sample spectra. The spectrum in Fig. 4a corresponds to the nucleus of the non-emission line galaxy II Zw 82. The spectrum is purely stellar, so it is straightforward to measure the pseudocontinuum. Figure 4b shows the nuclear spectrum of Haro 1, a star-forming galaxy. The continuum is also well defined. Overall, we find that the method works well for most spectra, with the exception of the nuclear regions of AGNs (see end of this Section).

The specific continuum wavelengths and corresponding fluxes for the non-ELGs and SFGs are shown in Table 4. The second line for each galaxy entry lists the errors with continuum measurements, calculated as the rms deviation from the average continuum flux.

4.2 Absorption line measurements

Figure 5: Log of the detection level of equivalent widths, i.e. log [ $EW/\sigma (EW)$], versus log EW for 4 absorption lines. The absorption line name, the number (SFG + non-ELG, EW > 1.5 Å) of plotted data and the 2$\sigma$, 3$\sigma$ detection levels (dashed lines) are indicated in each panels. The plotting symbols are coded according to spectral classification, the asterisks correspond to star-forming galaxies (SFGs), the triangles to non-ELGs.

 

Table 5: Wavelength windows used to measure the equivalent widths of the absorption lines.

No.	Window	Main
	(Å)	Absorber	Identification
1	3908-3952	CaII K	Ca K
2	3952-3988	CaII H+H$\epsilon$	Ca H
3	4082-4124	H$\delta$	H$\delta$
4	4150-4214	CN	CN
5	4284-4318	G band	G band
6	5156-5196	MgI+MgH	Mg
7	5880-5914	NaI	NaI

The spectrum of a galaxy is produced by the sum of the spectral characteristics of its stellar content (Weiss et al. 1995). Observations of the integrated spectra of galaxies can be used to determine the distribution in age and metal abundance of the stellar population in these systems and hence to determine their epoch of formation and subsequent star formation history (Arimoto & Yoshii 1986). To this aim, some strong, easily identifiable absorption features in our observed spectral range are measured, which include some age and metallicity sensitive absorption features. The absorption line names and adopted spectral wavelength windows are shown in Table 5.

The rest-frame equivalent widths of these absorption features were automatically computed by summing the observed fluxes below the continuum level, which itself is estimated by fitting a straight line to the fluxes in the above continuum regions. The equivalent widths of the absorption features for these non-ELGs and SFGs are presented in Table 6. The second line for each galaxy entry lists the errors on equivalent widths, which are computed from Eq. (2) in this paper. Figure 5 shows that logarithm of the detection level of EW versus log EW for 4 absorption lines, such as Ca K, Ca H, H$\delta$  and MgI+MgH when its EW > 1.5 Å. Our absorption feature EW limit is at a $3\sigma$ confidence level for the absorption lines of most galaxies. The mean uncertainty in these measurements is about 10% for those non-ELGs and about 15% for those SFGs.

4.3 4000 Å Balmer break index measurements

As well as the absorption features, we also measured the 4000 Å Balmer break. It is the strongest discontinuity in the optical spectrum of a galaxy and arises because of the accumulation of a large number of spectral lines in a narrow wavelength region (Bruzual 1983). The main contribution to the opacity comes from ionized metals. In hot stars, the elements are multiple ionized and the opacity decreases, so the 4000 Å break will be small for young stellar populations and large for old, metal-rich galaxies (Kauffmann et al. 2002).

We use the definition using narrower continuum bands than Bruzual's, which was introduced by Balogh et al (1999). The principal advantage of the narrow definition is that the index is considerably less sensitive to reddening effects. It is defined as the ratio of the average fluxes (for frequency unit) measured in the spectral ranges 4000-4100 Å and 3850-3950 Å: ${\rm D4000vn}=F_{\nu}$[4000-4100 Å]$/F_{\nu}$[3850-3950 Å]. The D4000vn index is simply a flux ratio and, hence the error is determined from standard propagation techniques. The D4000vn index value and its error for these non-ELGs and SFGs are presented in the last column of Table 6.

While for non-ELGs and SFGs the placement of the continuum is straightforward, this is not the case in the nuclear regions of most AGNs, where the numerous broad lines and intense non-stellar continuum complicate the analysis. The continuum points are impossible to determine accurately, and the equivalent widths of the absorption lines cannot be measured accurately. Therefore, for 10 AGNs, we only measured the integrated fluxes, F, and rest-frame equivalent widths, EWs, of the emission lines. We did not measure the continuum fluxes and equivalent widths of the absorption lines for those AGNs.