Results and discussion

Examples of blue and red spectra are shown in Figure 26.1. Emission-line fluxes were measured using the IRAF SPLOT software package. The presence of a conspicuous underlying stellar population, which is more evident in the blue spectra, in most observed regions complicates the measurements. An example of underlying absorption can be seen in the inset to the left panel of Figure 26.1. A two-component (emission and absorption) Gaussian fit was performed in order to correct

4,300

4,400

Wavelength (A)

Figure 26.2. An example of the fitting procedure used in order to correct the Balmer emission-line intensities for the presence of the underlying stellar population.

4,300

4,400

Wavelength (A)

Figure 26.2. An example of the fitting procedure used in order to correct the Balmer emission-line intensities for the presence of the underlying stellar population.

log R23

Figure 26.3. Left panel: the calibration of the R23 abundance parameter. Right panel: the calibration of the N2 abundance parameter. The location of CNSFRs is indicated.

log R23

Figure 26.3. Left panel: the calibration of the R23 abundance parameter. Right panel: the calibration of the N2 abundance parameter. The location of CNSFRs is indicated.

the Balmer lines for underlying absorption. An example of this procedure is shown in Figure 26.2.

The low excitation of the regions, as indicated by the weakness of the [H iii] 15007 A line (see the left panel of Figure 26.1), precludes the detection and measurement of the auroral [H iii] 13463 A necessary for the derivation of the electron temperature. It is therefore impossible to obtain a direct determination of the oxygen abundances. Empirical calibrations have to be used instead. In the left panel of Figure 26.3 we show the calibration of oxygen abundance by use of the commonly

Hii galaxies

Under-Solar-abundance H ii regions Over-Solar-metallicity H ii regions High-metallicity CNSFRs Low-metallicity CNSFRs

Hii galaxies

Under-Solar-abundance H ii regions Over-Solar-metallicity H ii regions High-metallicity CNSFRs Low-metallicity CNSFRs

Figure 26.4. Left panel: an empirical version of the N/O versus O/H relation. Right panel: the N2 abundance parameter as a function of excitation, as measured by the [S ii]/[S iii] ratio.

used R23 parameter defined as ([011] A.A.3727, 3729 + [O iii] XX4959, 5007)/H| (Pagel et al. 1979). Data on H ii galaxies, disk H ii regions, and CNSFRs are shown. The H ii-region sample (Perez-Montero & Diaz 2005) has been divided into under-Solar (open triangles) and over-Solar (filled triangles) according to the empirical criterion given by Diaz & Perez-Montero (2000), i.e. R23 < 0.47 and -0.5 < S23 < 0.28.1 The H ii-galaxy data (filled squares) come from Perez-Montero & Diaz (2003). The CNSFRs are represented by circles, solid ones for our observed objects and open ones for regions in NGC 3310 and NGC 7714 known to have under-Solar abundances (Pastoriza etal. 1993; Gonzalez-Delgado etal. 1995). As can be seen, the calibration is two-folded and has considerable scatter, and its high-abundance end is not well sampled. The positions of the observed CNSFRs are indicated. Their observed R23 values, lower than those of the lowest-abundance galaxy known (IZw18), indicate that CNSFRs belong to the high-abundance branch of the calibration, having possibly the highest metallicities shown by H ii-region-like objects. In the right panel of Figure 26.3 the positions of the regions are indicated in the N2 ([N ii]/Ha) abundance-calibration diagram (Denicolo et al. 2002), which reveals a linear behaviour. Again CNSFRs appear to have the highest oxygen abundances.

The left panel of Figure 26.4 shows the [N ii]/[0 ii] ratio versus the N2 abundance parameter. Since a good correlation has been found to exist between the [N ii]/[0 ii] ratio and the N+/0+ ionic abundance ratio, which in turn can be assumed to measure the N/0 ratio, this graph is the observational equivalent of the N/0 versus O/H diagram. We can see that a very tight correlation exists for all of the objects

1 The sulfur-abundance parameter S23 is defined as ([S ii] + [S iii])/H|.

Under-Solar-abundance Hii regions Over-Solar-abundance Hii regions High-metallicity CNSFRs H ii galaxies

Low-metallicity CNSFRs

Under-Solar-abundance Hii regions Over-Solar-abundance Hii regions High-metallicity CNSFRs H ii galaxies

Low-metallicity CNSFRs

Over-Solar-abundance Hii regions Under-Solar abundance Hii regions High-metallicity CNSFRs Hii galaxies

Low-metallicity CNSFRs y log h' = 0.

logh'

Figure 26.5. Left panel: the N2 versus n' relation. Right panel: the logarithmic relation between the [O ii]/[O iii] and [S ii]/[S iii] line ratios.

represented: H ii galaxies, low- and high-abundance H ii regions, and CNSFRs. Again our observed regions have the highest N/O ratios of the sample.

The right panel of Figure 26.4 shows the run of the degree of excitation with metallicity for the observed regions through the [S ii]/[S iii] ratio, which has been shown to be a good ionization-parameter indicator (Diaz et al. 1991), and the N2 parameter. It can be seen that the observed CNSFRs have the lowest excitation of the sample.

A hint on the ionizing temperature of the regions can be obtained through the use of the r¡' parameter, which is a measure of the softness of the ionizing radiation (Vílchez & Pagel 1988) and increases with decreasing ionizing temperature. The left panel of Figure 26.5 shows the run of r¡' with metallicity as parametrized by N2. Unexpectedly, CNSFRs have values of r¡' higher than those of over-Solar disk H ii regions. This is better appreciated in the right panel of the figure, where CNSFRs are seen to segregate from disk H ii regions in the [O ii]/[H iii] versus [S ii]/[S iii] diagram. The former cluster around the value of log r¡' = 0-0 (Tion ~ 40,000 K) while the latter cluster around log r¡' = 0-8 (T^n - 35,000 K).

Was this article helpful?

0 0

Post a comment