Abstract. We present a comprehensive hybrid library of synthetic stellar spectra based on three original grids of model atmosphere spectra by Kurucz (1995), Fluks et al. (1994), and Bessell et al. (1989, 1991, respectively. The combined library has been intended for multiple-purpose synthetic photometry applications and was constructed according to the precepts adopted by Buser & Kurucz (1992): (i) to cover the largest possible ranges in stellar parameters (T eff , log g, and [M/H]); (ii) to provide flux spectra with useful resolution on the uniform grid of wavelengths adopted by Kurucz (1995); and (iii) to provide synthetic broad-band colors which are highly realistic for the largest possible parameter and wavelength ranges.Because the most astrophysically relevant step consists in establishing a realistic library, the corresponding color calibration is described in some detail. Basically, for each value of the effective temperature and for each wavelength, we calculate the correction function that must be applied to a (theoretical) solar-abundance model flux spectrum in order for this to yield synthetic UBV RIJHKL colors matching the (empirical) color-temperature calibrations derived from observations. In this way, the most important systematic differences existing between the original model spectra and the observations can indeed be eliminated. On the other hand, synthetic UBV and Washington ultraviolet excesses δ (U−B) and δ (C−M) and δ (C−T1) obtained from the original giant and dwarf model spectra are in excellent accord with empirical metal-abundance calibrations (Lejeune & Buser 1996). Therefore, the calibration algorithm is designed in such a way as to preserve the original differential grid properties implied by metallicity and/or luminosity changes in the new library, if the above correction function for a solar-abundance model of a given effective temperature is also applied to models of the same temperature but different chemical compositions [M/H] and/or surface gravities log g. Send offprint requests to: R. BuserWhile the new library constitutes a first-order approximation to the program set out above, it will be allowed to develop toward the more ambitious goal of matching the full requirements imposed on a standard library. Major input for refinement and completion is expected from the extensive tests now being made in population and evolutionary synthesis studies of the integrated light of globular clusters (Lejeune 1997) and galaxies (Bruzual et al. 1997).

Abstract. A standard library of theoretical stellar spectra intended for multiple synthetic photometry applications including spectral evolutionary synthesis is presented. The grid includes M dwarf model spectra, hence complementing the first library version established in Paper I . It covers wide ranges of fundamental parameters: T eff : 50 000 K ∼ 2000 K, log g: 5.5 ∼ −1.02, and [M/H]: +1.0 ∼ −5.0. A correction procedure is also applied to the theoretical spectra in order to provide color-calibrated flux distributions over a large domain of effective temperatures. For this purpose, empirical T eff -color calibrations are constructed between 11500 K and 2000 K, and semi-empirical calibrations for non-solar abundances ([M/H] = −3.5 to +1.0) are established. Model colors and bolometric corrections for both the original and the corrected spectra, synthesized in the (UBV ) J (RI) C JHKLL M system, are given for the full range of stellar parameters. We find that the corrected spectra provide a more realistic representation of empirical stellar colors, though the method employed is not completely adapted to the lowest temperature models. In particular the original differential colors of the grid implied by metallicity and/or luminosity changes are not preserved below 2500 K. Limitations of the correction method used are also discussed.

Abstract. We extend the colour calibration of the widely used BaSeL standard stellar library (Lejeune et al. , 1998 The problem appears to be related to the longstanding incompleteness in our understanding of convection in late-type stellar evolution, but is also due to a serious lack of relevant observational calibration data that would help resolve, or at least further significant progress towards resolving this issue. In view of the most important applications of the BaSeL library, we here propose two different metallicity calibration versions: (1) the "WLBC 99" library, which consistently matches empirical colour-temperature relations and which, therefore, should make an ideal tool for the study of individual stars; and (2), the "PADOVA 2000" library, which provides isochrones from the Padova 2000 grid (Girardi et al. 2000) that successfully reproduce Galactic globular-cluster colour-absolute magnitude diagrams and which thus should prove particularly useful for studies of collective phenomena in stellar populations in clusters and galaxies.

A study of the blue compact dwarf (BCD) galaxy Mrk 996 based on high‐resolution optical Very Large Telescope Visible Multi‐Object Spectrograph integral field unit spectroscopy is presented. Mrk 996 displays multicomponent line emission, with most line profiles consisting of a narrow, central Gaussian [full width at half‐maximum (FWHM) ∼ 110 km s−1] with an underlying broad component (FWHM ∼ 400 km s−1). The broad H i Balmer component splits into two separate broad components inside a 1.5‐arcsec radius from the nucleus; these are attributed to a two‐armed minispiral. This spiral‐like nucleus rotates in the same sense as the extended narrow line ionized gas but is offset by ∼50 km s−1 from the systemic velocity of the galaxy. The rotation curve of Mrk 996 derived from the Hα narrow component yields a total mass of 5 × 108 M⊙ within a radius of 3 kpc. From the Hα luminosity we infer a global star formation rate of ∼2 M⊙ yr−1. The high excitation energy, high critical density [O iii]λ4363 and [N ii]λ5755 lines are only detected from the inner region and exist purely in broad component form, implying unusual excitation conditions. Surface brightness, radial velocity and FWHM maps for several emission components are presented. A separate physical analysis of the broad and narrow emission line regions is undertaken. We derive an upper limit of 10 000 K for the electron temperature of the narrow line gas, together with an electron density of 170 cm−3, typical of normal H ii regions. For the broad line component, measured [O iii] and [Fe iii] diagnostic line ratios are consistent with a temperature of 11 000 K and an electron density of 107 cm−3. The broad line emission regions show N/H and N/O enrichment factors of ∼20 relative to the narrow line regions, but no He/H, O/H, S/H or Ar/H enrichment is inferred. Previous studies indicated that Mrk 996 showed anomalously high N/O ratios compared with BCDs of a similar metallicity. Our multicomponent analysis yields a revised metallicity of ≥0.5 Z⊙ (12 + log O/H = 8.37) for both the narrow and broad gas components, significantly higher than previous studies. As a result the narrow line region's N/O ratio is now typical for the galaxy's metallicity. The narrow line component's N/O ratio peaks outside the core region, spatially correlating with ∼3‐Myr‐old stellar populations. The dominant line excitation mechanism is photoionization by the ∼3000 Wolf–Rayet stars and ∼150 000 O‐type stars estimated to be present in the core. This is indeed a peculiar BCD, with extremely dense zones of gas in the core, through which stellar outflows and possible shock fronts permeate contributing to the excitation of the broad line emission.