Stellar indices and kinematics in Seyfert 1 nuclei

We present spectra of six type 1 and two type 2 Seyfert galaxies, a starburst galaxy and a compact narrow-line radio galaxy, taken in two spectral ranges centred around the near-infrared Ca  ii triplet (∼8600 Å), and the Mgb stellar feature at 5180 Å. We measured the equivalent widths (EWs) of these features and the Fe₅₂ and Fe₅₃ spectral indices.
We found that the strength of the infrared Ca  ii triplet (CaT) in type 1 Seyfert galaxies with prominent central point sources is larger than what would be expected from the observed strength of the blue indices. This could be explained by the presence of red supergiants in the nuclei of Seyfert 1 galaxies. On the other hand, the blue indices of these galaxies could also be diluted by the strong Fe  ii multiplets that can be seen in their spectra.
We have also measured the stellar‐ and gas-velocity dispersions of the galaxies in the sample. The stellar velocity dispersions were measured using both the Mgb and CaT stellar features. The velocity dispersion of the gas in the narrow-line region (NLR) was measured using the strong emission lines [O  iii ] λλ 5007, 4959 and [S  iii ] λ 9069. We compare the gas- and star-velocity dispersions and find that the magnitudes of both are correlated in Seyfert galaxies.
Most of the Seyfert 1 galaxies that we observe have stellar‐velocity dispersions somewhat greater than that of the gas in the NLR.     

The CaT strength in Seyfert nuclei revisited: analysing young stars and non-stellar light contributions to the spectra

In a former paper, we have presented spectra of 64 active, nine normal and five starburst galaxies in the region around the near-infrared calcium triplet (CaT) absorption lines and the [S  iii ]λ9069 line. In the present paper, we analyse the CaT strength ( W _CaT) and kinematical products derived in that study, namely stellar  (σ_★)  and ionized gas (σ_gas) velocity dispersions. Our main results may be summarized as follows. (1) Type 2 Seyfert galaxies show no sign of dilution in W _CaT with respect to the values spanned by normal galaxies, even when optical absorption lines such as the Ca  ii K band at 3933 Å are much weaker than in old, bulge-like stellar populations. (2) The location of type 2 Seyfert galaxies in the   W _CaT– W _CaK  plane is consistent with evolutionary synthesis models. The implication is that the source responsible for the dilution of optical lines in these active galactic nuclei (AGN) is a young stellar population, rather than an AGN featureless continuum, confirming the conclusion of the pioneer study of Terlevich, Díaz & Terlevich. (3) In type 1 Seyfert galaxies, both   W _{[S  iii ]}  and W _CaT tend to be diluted due to the presence of a non-stellar component, in agreement with the unification paradigm. (4) A comparison of  σ_★  with σ_gas (obtained from the core of the [S  iii ] emitting line) confirms the existence of a correlation between the typical velocities of stars and clouds of the narrow line region. The strength and scatter around this correlation are similar to those previously obtained from the [O  iii ]λ5007 linewidth.     

Towards a precision cosmology from starburst galaxies at z > 2

This work investigates the use of a well-known empirical correlation between the velocity dispersion, metallicity and luminosity in Hβ of nearby H  ii galaxies to measure the distances to H  ii -like starburst galaxies at high redshifts. This correlation is applied to a sample of 15 starburst galaxies with redshifts between   z = 2.17  and   z = 3.39  to constrain  Ω_m  , using data available from the literature. A best-fitting value of  Ω_m= 0.21^+0.30_−0.12  in a Λ-dominated universe and of  Ω_m= 0.11^+0.37_−0.19  in an open universe is obtained. A detailed analysis of systematic errors, their causes and their effects on the values derived for the distance moduli and  Ω_m  is carried out. A discussion of how future work will improve constraints on  Ω_m  by reducing the errors is also presented.     

The Hα luminosity function and star formation rate up to z ∼ 1

We describe ISAAC/ESO-VLT observations of the Hαλ6563 Balmer line of 33 field galaxies from the Canada–France Redshift Survey (CFRS) with redshifts selected between 0.5 and 1.1. We detect Hα in emission in 30 galaxies and compare the properties of this sample with the low-redshift sample of CFRS galaxies at   z ∼ 0.2  . We find that the Hα luminosity,   L (Hα)  , is tightly correlated to   M ( B _AB)  in the same way for both the low- and high-redshift samples.   L (Hα)  is also correlated to L ([O  ii ]λ3727), and again the relation appears to be similar at low and high redshifts. The ratio L (lsqb;O  ii ])/   L (Hα)  decreases for brighter galaxies by as much as a factor of 2 on average. Derived from the Hα luminosity function, the comoving Hα luminosity density increases by a factor 12 from  〈 z 〉= 0.2  to  〈 z 〉= 1.3  . Our results confirm a strong rise of the star formation rate (SFR) at   z < 1.3  , proportional to  (1 + z )^4.1±0.3  (with   H ₀= 50 km s⁻¹ Mpc⁻¹, q ₀= 0.5  ). We find an average  SFR(2800 Å)/SFR (Hα)  ratio of 3.2 using the Kennicutt SFR transformations. This corresponds to the dust correction that is required to make the near-ultraviolet data consistent with the reddening-corrected Hα data within the self-contained, I -selected CFRS sample.     

The properties of 70 μm-selected high-redshift galaxies in the Extended Groth Strip

We examine the infrared properties of 43 high-redshift (0.1 < z < 1.2), infrared-luminous galaxies in the Extended Groth Strip (EGS), selected by a deep 70 μm survey with the Multiband Imaging Photometer on Spitzer (MIPS). In addition and with reference to starburst-type spectral energy distributions (SEDs), we derive a set of equations for estimating the total infrared luminosity ( L _IR) in the range 8–1000 μm using photometry from at least one MIPS band. 42 out of 43 of our sources' optical/infrared SEDs (λ_observed < 160 μm) are starburst type, with only one object displaying a prominent power-law near-infrared continuum. For a quantitative analysis, models of radiation transfer in dusty media are fit on to the infrared photometry, revealing that the majority of galaxies are represented by high extinction, A _v > 35, and for a large fraction (∼50 per cent) the SED turns over into the Rayleigh–Jeans regime at wavelengths longward of 90 μm. For comparison, we also fit semi-empirical templates based on local galaxy data; however, these underestimate the far-infrared SED shape by a factor of at least 2 and in extreme cases up to 10 for the majority (∼70 per cent) of the sources. Further investigation of SED characteristics reveals that the mid-infrared (70/24 μm) continuum slope is decoupled from various galaxy properties such as the total infrared luminosity and far-infrared peak, quantified by the L ₁₆₀/ L ₇₀ ratio. In view of these results, we propose that these high-redshift galaxies have different properties to their local counterparts, in the sense that large amounts of dust cause heavy obscuration and are responsible for an additional cold emissive component, appearing as a far-infrared excess in their SEDs.     

Hα photometry of Abell 2390

We present the results of a search for strong H α emission line galaxies (rest frame equivalent widths greater than 50 Å) in the z ≈0.23 cluster Abell 2390. The survey contains 1189 galaxies over 270 arcmin², and is 50 per cent complete at M _r ≈−17.5+5 log  h . The fraction of galaxies in which H α is detected at the 2 σ level rises from 0.0 in the central regions (excluding the cD galaxy) to 12.5±8 per cent at R ₂₀₀. For 165 of the galaxies in our catalogue, we compare the H α equivalent widths with their [O  ii ] λ 3727 equivalent widths, from the Canadian Network for Observational Cosmology (CNOC1) spectra. The fraction of strong H α emission line galaxies is consistent with the fraction of strong [O  ii ] emission galaxies in the CNOC1 sample: only 2±1 per cent have no detectable [O  ii ] emission and yet significant (>2 σ ) H α equivalent widths. Dust obscuration, non-thermal ionization, and aperture effects are all likely to contribute to this non-correspondence of emission lines. We identify six spectroscopically 'secure' k+a galaxies [ W ₀(O  ii )<5 Å and W ₀(H δ )≳5 Å]; at least two of these show strong signs in H α of star formation in regions that are covered by the slit from which the spectra were obtained. Thus, some fraction of galaxies classified as k+a based on spectra shortward of 6000 Å are likely to be undergoing significant star formation. These results are consistent with a 'strangulation' model for cluster galaxy evolution, in which star formation in cluster galaxies is gradually decreased, and is neither enhanced nor abruptly terminated by the cluster environment.     

Molecular hydrogen emission in Cygnus A

We present J , H and K -band spectroscopy of Cygnus A, spanning 1.0–2.4 μm in the rest-frame and hence several rovibrational H₂, H recombination and [Fe  ii ] emission lines. The lines are spatially extended by up to 6 kpc from the nucleus, but their distinct kinematics indicate that the three groups (H, H₂ and [Fe  ii ]) are not wholly produced in the same gas. The broadest line, [Fe  ii ] λ 1.644, exhibits a non-Gaussian profile with a broad base (FWHM≃1040 km s⁻¹), perhaps because of the interaction with the radio source. Extinctions to the line-emitting regions substantially exceed earlier measurements based on optical H recombination lines.
Hard X-rays from the quasar nucleus are likely to dominate the excitation of the H₂ emission. The results of Maloney, Hollenbach & Tielens are thus used to infer the total mass of gas in H₂ v=1–0 S(1)-emitting clouds as a function of radius, for gas densities of 10³ and 10⁵ cm⁻³, and stopping column densities N _H=10²²–10²⁴ cm⁻². Assuming azimuthal symmetry, at least 2.3×10⁸ M_⊙ of such material is present within 5 kpc of the nucleus, if the line-emitting clouds see an unobscured quasar spectrum. Alternatively, if the bulk of the X-ray absorption to the nucleus inferred by Ueno et al. actually arises in a circumnuclear torus, the implied gas mass rises to ∼10¹⁰ M_⊙. The latter plausibly accounts for 10⁹ yr of mass deposition from the cluster cooling flow, for which within this radius.     

Near-infrared properties of i-drop galaxies in the Hubble Ultra Deep Field

We analyse near-infrared Hubble Space Telescope ( HST )/Near-Infrared Camera and Multi-Object Spectrometer F 110 W ( J ) and F 160 W ( H ) band photometry of a sample of 27 i '-drop candidate   z ≃ 6  galaxies in the central region of the HST /Advanced Camera for Surveys Ultra Deep Field . The infrared colours of the 20 objects not affected by near neighbours are consistent with a high-redshift interpretation. This suggests that the low-redshift contamination of this i '-drop sample is smaller than that observed at brighter magnitudes, where values of 10–40 per cent have been reported. The J – H colours are consistent with a slope flat in   f_ν ( f_λ ∝λ⁻²)  , as would be expected for an unreddened starburst. However, there is evidence for a marginally bluer spectral slope  ( f_λ ∝λ^−2.2)  , which is perhaps indicative of an extremely young starburst (∼10 Myr old) or a top heavy initial mass function and little dust. The low levels of contamination, median photometric redshift of   z ∼ 6.0  and blue spectral slope, inferred using the near-infrared data, support the validity of the assumptions in our earlier work in estimating the star formation rates, and that the majority of the i -drop candidates galaxies lie at   z ∼ 6  .     

Spectroscopy of the optical Einstein ring 0047 – 2808

We present optical and near-infrared spectroscopic observations of the optical Einstein ring 0047 – 2808. We detect both [O III ] lines λλ4959, 5007 near ∼ 2.3 μm, confirming the redshift of the lensed source as z  = 3.595. The Lyα line is redshifted relative to the [O III ] line by 140 ± 20 km s⁻¹. Similar velocity shifts have been seen in nearby starburst galaxies. The [O III ] line is very narrow, 130 km s⁻¹ FWHM. If the ring is the image of the centre of a galaxy, the one-dimensional stellar velocity dispersion σ = 55 km s⁻¹ is considerably smaller than the value predicted by Baugh et al. for the somewhat brighter Lyman-break galaxies. The Lyα line is significantly broader than the [O III ] line, probably due to resonant scattering. The stellar central velocity dispersion of the early-type deflector galaxy at z  = 0.485 is 250 ± 30 km s⁻¹. This value is in good agreement both with the value predicted from the radius of the Einstein ring (and a singular isothermal sphere model for the deflector), and with the value estimated from the D _n −σ relation.     

The anomalous properties of Markarian 1460

We present and discuss optical, near-infrared and H  i measurements of the galaxy Markarian 1460 at a distance of 19 Mpc in the Ursa Major Cluster. This low-luminosity ( M _B =−14) galaxy is unusual because (i) it is blue ( B − R =0.8) and has the spectrum of an H  ii galaxy, (ii) it has a light profile that is smooth and well fitted by an r ^1/4 and not an exponential function at all radii larger than the seeing, and (iii) it has an observed central brightness of about μ _B =20 mag arcsec⁻² , intermediate between those of elliptical galaxies (on the bright μ _B side) and normal low-luminosity dwarf irregular (on the low μ _B side) galaxies. No other known galaxy exhibits all these properties in conjunction. On morphological grounds this galaxy looks like a normal distant luminous elliptical galaxy, since the Fundamental Plane tells us that higher luminosity normal elliptical galaxies tend to have lower surface-brightnesses. Markarian 1460 has 2×10⁷ M_⊙ of H  i and a ratio M (H  i )/ L _B of 0.2, which is low compared to the typical values for star-forming dwarf galaxies. From the high surface-brightness and r ^1/4 profile, we infer that the baryonic component of Markarian 1460 has become self-gravitating through dissipative processes. From the colours, radio continuum, H  i and optical emission line properties, and yet smooth texture, we infer that Markarian 1460 has had significant star formation as recently as ∼1 Gyr ago but not today.     

Stellar velocity dispersion in narrow-line Seyfert 1 galaxies

Several authors have recently explored, for narrow-line Seyfert 1 galaxies (NLS1s), the relationship between black hole mass ( M _BH) and stellar velocity dispersion (σ_*). Their results are more or less in agreement and seem to indicate that NLS1s fill the region below the fit obtained by Tremaine et al., showing a range of σ_* similar to that of Seyfert 1 galaxies, and a lower M _BH. Until now, the [O  iii ] width has been used in place of the stellar velocity dispersion, but some indications have begun to arise against the effectiveness of the gaseous kinematics in representing the bulge potential, at least in NLS1s. Bian & Zhao have stressed the urgency of producing true σ_* measurements. Here, we present new stellar velocity dispersions obtained through direct measurements of the Ca  ii absorption triplet (∼8550 Å) in the nuclei of eight NLS1 galaxies. The resulting σ_* values and a comparison with σ_[O III] confirm our suspicion that [O  iii ] typically overestimates the stellar velocity dispersion. We demonstrate that NLS1s follow the   M _BH–σ_*  relation as Seyfert 1, quasars and non-active galaxies.     

X-ray observations of the edge-on star-forming galaxy NGC 891 and its supernova SN1986J

We present XMM–Newton observations of NGC 891, a nearby edge-on spiral galaxy. We analyse the extent of the diffuse emission emitted from the disc of the galaxy, and find that it has a single-temperature profile with best-fitting temperature of 0.26 keV, though the fit of a dual-temperature plasma with temperatures of 0.08 and 0.30 keV is also acceptable. There is a considerable amount of diffuse X-ray emission protruding from the disc in the north-west direction out to approximately 6 kpc. We analyse the point-source population using a Chandra observation, using a maximum-likelihood method to find that the slope of the cumulative luminosity function of point sources in the galaxy is  −0.77^+0.13_−0.1  . Using a sample of other local galaxies, we compare the X-ray and infrared properties of NGC 891 with those of 'normal' and starburst spiral galaxies, and conclude that NGC 891 is most likely a starburst galaxy in a quiescent state. We establish that the diffuse X-ray luminosity of spirals scales with the far-infrared luminosity as   L _X∝ L ^0.87±0.07_FIR  , except for extreme starbursts, and NGC 891 does not fall in the latter category. We study the supernova SN1986J in both XMM–Newton and Chandra observations, and find that the X-ray luminosity has been declining with time more steeply than expected  ( L _X∝ t ⁻³)  .     

Keck spectroscopy of the faint dwarf elliptical galaxy population in the Perseus Cluster core: mixed stellar populations and a flat luminosity function

We present the result of a photometric and Keck low-resolution imaging spectrometer (LRIS) spectroscopic study of dwarf galaxies in the core of the Perseus Cluster, down to a magnitude of   M _B =−12.5  . Spectra were obtained for 23 dwarf-galaxy candidates, from which we measure radial velocities and stellar population characteristics from absorption line indices. From radial velocities obtained using these spectra, we confirm 12 systems as cluster members, with the remaining 11 as non-members. Using these newly confirmed cluster members, we are able to extend the confirmed colour–magnitude relation for the Perseus Cluster down to   M _B =−12.5  . We confirm an increase in the scatter about the colour–magnitude relationship below   M _B =−15.5  , but reject the hypothesis that very red dwarfs are cluster members. We measure the faint-end slope of the luminosity function between   M _B =−18  and −12.5, finding  α=−1.26 ± 0.06  , which is similar to that of the field. This implies that an overabundance of dwarf galaxies does not exist in the core of the Perseus Cluster. By comparing metal and Balmer absorption line indices with α-enhanced single stellar population models, we derive ages and metallicities for these newly confirmed cluster members. We find two distinct dwarf elliptical populations: an old, metal-poor population with ages ∼8 Gyr and metallicities  [Fe/H] < −0.33  , and a young, metal-rich population with ages <5 Gyr and metallicities  [Fe/H] > −0.33  . Dwarf galaxies in the Perseus Cluster are therefore not a simple homogeneous population, but rather exhibit a range in age and metallicity.     

On the relation between electron temperatures in the O+ and O++ zones in high-metallicity H ii regions

We suggest a new way to establish the relation between the electron temperature t ₃ within the [O  iii ] zone and the electron temperature t ₂ within the [O  ii ] zone in high-metallicity  (12 + log(O/H) > 8.25)  H  ii regions. The   t ₂– t ₃  diagram is constructed by applying our method to a sample of 372 H  ii regions. We find that the correlation between t ₂ and t ₃ is tight and can be approximated by a linear expression. The new   t ₂– t ₃  relation can be used to determine t ₂ and accurate abundances in high-metallicity H  ii regions with a measured t ₃. It can also be used in conjunction with the ff relation for the determination of t ₃ and t ₂ and oxygen abundances in high-metallicity H  ii regions, where the [O  iii ]λ4363 auroral line is not detected. The derived   t ₂– t ₃  relation is independent of photoionization models of H  ii regions.     

Constraints on the nuclear emission of the Circinus galaxy: the torus

In the context of the unified model of Seyfert galaxies, we use observations from the literature and a radiative transfer model to investigate the near-IR to mm emission produced by the presumed torus in the Circinus galaxy, from 2 μm to 1.3 mm. From the infrared SED modelling, we find that the total luminosity ( L _IR) in this wavelength range consists of similar contributions from the torus and starburst with a ratio of nuclear luminosity to starburst luminosity ( L _NUC/ L _SB)∼0.8 .
By using a similar torus model to that of NGC 1068, but without the conical dust , we find an upper limit to the outer torus radius of ∼12 pc with a best fit of ∼2 pc. The upper limit torus size estimated from the radiative transfer modelling is consistent with the 16-pc torus radius estimated from near-IR imaging polarimetry of Circinus.     

A constant dark matter halo surface density in galaxies

We confirm and extend the recent finding that the central surface density  μ_0D≡ r ₀ρ₀  of galaxy dark matter haloes, where r ₀ and  ρ₀  are the halo core radius and central density, is nearly constant and independent of galaxy luminosity. Based on the co-added rotation curves (RCs) of ∼1000 spiral galaxies, the mass models of individual dwarf irregular and spiral galaxies of late and early types with high-quality RCs, and the galaxy–galaxy weak-lensing signals from a sample of spiral and elliptical galaxies, we find that  log μ_0D= 2.15 ± 0.2  in units of  log(M_⊙ pc⁻²)  . We also show that the observed kinematics of Local Group dwarf spheroidal galaxies are consistent with this value. Our results are obtained for galactic systems spanning over 14 mag, belonging to different Hubble types and whose mass profiles have been determined by several independent methods. In the same objects, the approximate constancy of  μ_0D  is in sharp contrast to the systematical variations, by several orders of magnitude, of galaxy properties, including  ρ₀  and central stellar surface density.     

On the H i content, dust-to-gas ratio and nature of Mg ii absorbers

We estimate the mean dust-to-gas ratio of Mg  ii absorbers as a function of rest equivalent width W ₀ and redshift over the range  0.5 < z < 1.4  . Using the expanded Sloan Digital Sky Survey/ Hubble Space Telescope sample of low-redshift Lyman-α absorbers, we first show the existence of a 8σ correlation between the mean hydrogen column density  〈 N _{H  i} 〉  and W ₀, an indicator of gas velocity dispersion. By combining these results with recent dust-reddening measurements, we show that the mean dust-to-gas ratio of Mg  ii absorbers does not appreciably depend on rest equivalent width. Assuming that, on average, dust-to-gas ratio is proportional to metallicity, we find its redshift evolution to be consistent with that of   L ^★  galaxies from   z = 0.5  to 1.4, and we show that our constraints disfavour dwarf galaxies as the origin of such absorbers. We discuss other scenarios and favour galactic outflows from  ∼ L ^★  galaxies as the origin of the majority of strong Mg  ii absorbers. Finally, we show that, once evolutionary effects are taken into account, the Bohlin et al. relation between A _V and N _H is also satisfied by strong Mg  ii systems down to lower column densities than those probed in our Galaxy.     

The quasar fraction in low-frequency-selected complete samples and implications for unified schemes

Low-frequency radio surveys are ideal for selecting orientation-independent samples of extragalactic sources because the sample members are selected by virtue of their isotropic steep-spectrum extended emission. We use the new 7C Redshift Survey along with the brighter 3CRR and 6C samples to investigate the fraction of objects with observed broad emission lines – the 'quasar fraction'– as a function of redshift and of radio and narrow-emission-line luminosity. We find that the quasar fraction is more strongly dependent upon luminosity (both narrow-line and radio) than it is on redshift. Above a narrow [O  ii ] emission-line luminosity of log₁₀( L _{[O  ii ]}/W)≳35 [or radio luminosity log₁₀( L ₁₅₁/W Hz⁻¹ sr⁻¹)≳ 26.5], the quasar fraction is virtually independent of redshift and luminosity; this is consistent with a simple unified scheme with an obscuring torus with a half-opening angle θ _trans≈53°. For objects with less luminous narrow lines, the quasar fraction is lower. We show that this is not due to the difficulty of detecting lower luminosity broad emission lines in a less luminous, but otherwise similar, quasar population. We discuss evidence which supports at least two probable physical causes for the drop in quasar fraction at low luminosity: (i) a gradual decrease in θ _trans and/or a gradual increase in the fraction of lightly reddened (0≲ A _V ≲5) lines of sight with decreasing quasar luminosity; and (ii) the emergence of a distinct second population of low-luminosity radio sources which, like M87, lack a well-fed quasar nucleus and may well lack a thick obscuring torus.     

Near-infrared study of CIZA J1324.7-5736, the second richest cluster of galaxies in the Great Attractor

We present the result of a deep near-infrared survey of the newly identified X-ray luminous cluster of galaxies CIZA J1324.7-5736 in the Great Attractor (GA) region. In a 35 × 35 arcmin² region, 111 galaxy candidates with   r _{Ks 20} > arcsec  are identified. Comparison of the extinction-corrected   K _s   -band luminosity function of CIZA J1324.7-5736 with those of nearby clusters indicates that the richness class of CIZA J1324.7-5736 is almost the same as, or richer than, the Pavo, Centaurus and Hydra clusters but poorer than the Coma, Perseus and Norma clusters. CIZA J1324.7-5736 is possibly the second richest cluster in the GA region following the Norma cluster. The position of CIZA J1324.7-5736 [in the ( l , b , v ) space] is close to the Centaurus–Crux cluster and the agglomeration of galaxies detected by the Parkes H  i survey. CIZA J1324.7-5736, together with the Centaurus–Crux cluster and the H  i galaxy agglomeration, is most likely to be one of the richest local concentrations in the GA overdensity of galaxies.     

New light on the search for low-metallicity galaxies – I. The N2 calibrator

We present a simple metallicity estimator based on the logarithmic [N  ii ]   λ 6584/H α   ratio, hereafter N2, which we envisage will become very useful for ranking galaxies in a metallicity sequence from redshift survey-quality data even for moderately low spectral resolution.
We have calibrated the N2 estimator using a compilation of H  ii galaxies having accurate oxygen abundances, plus photoionization models covering a wide range of abundances. The comparison of models and observations indicates that both primary and secondary nitrogen are important for the relevant range of metallicities.
The N2 estimator follows a linear relation with log(O/H) that holds for the whole abundance range covered by the sample, from approximately  1/50th  to twice the Solar value  [7.2<12+log(O/H)<9.1]  . We suggest that the ([S  ii ]   λλ 6717,6731/H α )  ratio (hereafter S2) can also be used as a rough metallicity indicator. Because of its large scatter the S2 estimator will be useful only in systems with very low metallicity, where [N  ii ] λ 6584 is not detected or in low-resolution spectra where [N  ii ] λ 6584 is blended with H α .