Stellar Population Synthesis Models Research Articles

A comprehensive comparative test of seven widely used spectral synthesis models against multi-band photometry of young massive-star clusters

We test the predictions of spectral synthesis models based on seven different massive-star prescriptions against Legacy ExtraGalactic UV Survey (LEGUS) observations of eight young massive clusters in two local galaxies, NGC 1566 and NGC 5253, chosen because predictions of all seven models are available at the published galactic metallicities. The high angular resolution, extensive cluster inventory and full near-ultraviolet to near-infrared photometric coverage make the LEGUS dataset excellent for this study. We account for both stellar and nebular emission in the models and try two different prescriptions for attenuation by dust. From Bayesian fits of model libraries to the observations, we find remarkably low dispersion in the median E(B-V) (~0.03 mag), stellar masses (~10^4 M_\odot) and ages (~1 Myr) derived for individual clusters using different models, although maximum discrepancies in these quantities can reach 0.09 mag and factors of 2.8 and 2.5, respectively. This is for ranges in median properties of 0.05-0.54 mag, 1.8-10x10^4 M_\odot and 1.6-40 Myr spanned by the clusters in our sample. In terms of best fit, the observations are slightly better reproduced by models with interacting binaries and least well reproduced by models with single rotating stars. Our study provides a first quantitative estimate of the accuracies and uncertainties of the most recent spectral synthesis models of young stellar populations, demonstrates the good progress of models in fitting high-quality observations, and highlights the needs for a larger cluster sample and more extensive tests of the model parameter space.

Breaking the disc-halo degeneracy in NGC 1291 using hydrodynamic simulations

AbstractWe present a pilot study on the nearby massive galaxy NGC 1291, in which we aim to constrain the dark matter in the inner regions, by obtaining a dynamical determination of the disc mass-to-light ratio (M/L). To this aim, we model the bar-induced dust lanes in the galaxy, using hydrodynamic gas response simulations. The models have three free parameters, the M/L of the disc, the bar pattern speed and the disc height function. We explore the parameter space to find the best fit models, i.e. those in which the morphology of the shocks in the gas simulations matches the observed dust lanes. The best-fit models suggest that the M/L of NGC 1291 agrees with that predicted by stellar population synthesis models in the near-infrared (≈ 0.6M⊙/L⊙), which leads to a borderline maximum disc for this galaxy. The bar rotates fast, with corotation radius ⩽ 1.4 times the bar length. Additionally, we find that the height function has a significant effect on the results, and can bias them towards lower or higher M/L.

The Lyman-continuum photon production efficiency in the high-redshift Universe

Abstract The Lyman-continuum photon production efficiency (ξion) is a critical ingredient for inferring the number of photons available to reionize the intergalactic medium. To estimate the theoretical production efficiency in the high-redshift Universe we couple the BlueTides cosmological hydrodynamical simulation with a range of stellar population synthesis models. We find Lyman-continuum photon production efficiencies of log10(ξion/erg−1 Hz) ≈ 25.1–25.5 depending on the choice of stellar population synthesis model. These results are broadly consistent with recent observational constraints at high-redshift though favour a model incorporating the effects of binary evolution.

Stellar population constraints on the ages of galactic bars

We present a study of the stellar populations within the central regions of four nearby barred galaxies, and use a novel technique to constrain the duration of bar activity. We focus on the star formation 'desert', a region within each of these galaxies where star formation appears to have been suppressed by the bar. New H beta spectroscopic data are presented, and used to produce spectroscopic line indices which are compared with theoretical predictions from population synthesis models for simple stellar populations and temporally truncated star formation histories. This analysis shows that the dearth of star formation activity in these regions appears to have been continuing for at least 1 Gyr, with timescales of several Gyr indicated for two of the galaxies. This favours models in which strong bars can be long-lived features of galaxies, but our results also indicate a significant diversity in stellar population ages, and hence in the implied histories of bar activity in these four galaxies.

Quantifying correlations between galaxy emission lines and stellar continua

We analyse the correlations between continuum properties and emission line equivalent widths of star-forming and active galaxies from the Sloan Digital Sky Survey. Since upcoming large sky surveys will make broad-band observations only, including strong emission lines into theoretical modelling of spectra will be essential to estimate physical properties of photometric galaxies. We show that emission line equivalent widths can be fairly well reconstructed from the stellar continuum using local multiple linear regression in the continuum principal component analysis (PCA) space. Line reconstruction is good for star-forming galaxies and reasonable for galaxies with active nuclei. We propose a practical method to combine stellar population synthesis models with empirical modelling of emission lines. The technique will help generate more accurate model spectra and mock catalogues of galaxies to fit observations of the new surveys. More accurate modelling of emission lines is also expected to improve template-based photometric redshift estimation methods. We also show that, by combining PCA coefficients from the pure continuum and the emission lines, automatic distinction between hosts of weak active galactic nuclei (AGNs) and quiescent star-forming galaxies can be made. The classification method is based on a training set consisting of high-confidence starburst galaxies and AGNs, and allows for the similar separation of active and star-forming galaxies as the empirical curve found by Kauffmann et al. We demonstrate the use of three important machine learning algorithms in the paper: k-nearest neighbour finding, k-means clustering and support vector machines.

Sulphur abundance determinations in star-forming regions – I. Ionization correction factor

In the present work we used a grid of photoionization models combined with stellar population synthesis models to derive reliable Ionization Correction Factors (ICFs) for the sulphur in star-forming regions. These models cover a large range of nebular parameters and yielding ionic abundances in consonance with those derived through optical and infrared observational data of star-forming regions. From our theoretical ICFs, we suggested an {\alpha} value of 3.27 in the classical Stasinska formulae. We compared the total sulphur abundance in the gas phase of a large sample of objects by using our Theoretical ICF and other approaches. In average, the differences between the determinations via the use of the different ICFs considered are similar to the uncertainties in the S/H estimations. Nevertheless, we noted that for some objects it could reach up to about 0.3 dex for the low metallicity regime. Despite of the large scatter of the points, we found a trend of S/O ratio to decrease with the metallicity, independently of the ICF used to compute the sulphur total abundance.

Dynamical measurement of the stellar surface density of face-on galaxies

The DiskMass survey recently provided measurements of the vertical velocity dispersions of disk stars in a sample of nearly face-on galaxies. By setting the disk scale-heights to be equal to those of edge-on galaxies with similar scale-lengths, it was found that these disks must be sub-maximal, with surprisingly low K-band mass-to-light ratios of the order of $M_\star/L_K \simeq 0.3 M_\odot/L_\odot$. This study made use of a simple relation between the disk surface density and the measured velocity dispersion and scale height of the disk, neglecting the shape of the rotation curve and the dark matter contribution to the vertical force, which can be especially important in the case of sub-maximal disks. Here, we point out that these simplifying assumptions led to an overestimation of the stellar mass-to-light ratios. Relaxing these assumptions, we compute even lower values than previously reported for the mass-to-light ratios, with a median $M_\star/L_K \simeq 0.18 M_\odot/L_\odot$, where 14 galaxies have $M_\star/L_K < 0.11$. Invoking prolate dark matter halos made only a small difference to the derived $M_\star/L_K$, although extreme prolate halos ($q>1.5$ for the axis ratios of the potential) might help. The cross-terms in the Jeans equation are also generally negligible. These deduced K-band stellar mass-to-light ratios are even more difficult to reconcile with stellar population synthesis models than the previously reported ones.

Biases and systematics in the observational derivation of galaxy properties: comparing different techniques on synthetic observations of simulated galaxies

We study the sources of biases and systematics in the derivation of galaxy properties of observational studies, focusing on stellar masses, star formation rates, gas/stellar metallicities, stellar ages and magnitudes/colors. We use hydrodynamical cosmological simulations of galaxy formation, for which the real quantities are known, and apply observational techniques to derive the observables. We also make an analysis of biases that are relevant for a proper comparison between simulations and observations. For our study, we post-process the simulation outputs to calculate the galaxies' spectral energy distributions (SEDs) using Stellar Population Synthesis models and also generating the fully-consistent far UV-submillimeter wavelength SEDs with the radiative transfer code SUNRISE. We compared the direct results of simulations with the observationally-derived quantities obtained in various ways, and found that systematic differences in all studied galaxy properties appear, which are caused by: (1) purely observational biases, (2) the use of mass-weighted/luminosity-weighted quantities, with preferential sampling of more massive/luminous regions, (3) the different ways to construct the template of models when a fit to the spectra is performed, and (4) variations due to the use of different calibrations, most notably in the cases of the gas metallicities and star formation rates. Our results show that large differences can appear depending on the technique used to derive galaxy properties. Understanding these differences is of primary importance both for simulators, to allow a better judgement on similarities/differences with observations, and for observers, to allow a proper interpretation of the data.

Single stellar populations in the near-infrared

We present unresolved single stellar population synthesis models in the near-infrared (NIR) range. The extension to the NIR is important for the study of early-type galaxies, since these galaxies are predominantly old and therefore emit most of their light in this wavelength range. The models are based on a library of empirical stellar spectra, the NASA infrared telescope facility (IRTF) spectral library. Integrating these spectra along theoretical isochrones, while assuming an initial mass function (IMF), we have produced model spectra of single age-metallicity stellar populations at a resolution R~2000. These models can be used to fit observed spectral of globular clusters and galaxies, to derive their age distribution, chemical abundances and IMF. The models have been tested by comparing them to observed colours of elliptical galaxies and clusters in the Magellanic Clouds. Predicted absorption line indices have been compared to published indices of other elliptical galaxies. The comparisons show that our models are well suited for studying stellar populations in unresolved galaxies. They are particularly useful for studying the old and intermediate-age stellar populations in galaxies, relatively free from contamination of young stars and extinction by dust. These models will be indispensable for the study of the upcoming data from JWST and extremely large telescopes, such as the E-ELT.

The Local Group as a time machine: studying the high-redshift Universe with nearby galaxies

We infer the UV luminosities of Local Group galaxies at early cosmic times (z ∼ 2 and z ∼ 7) by combining stellar population synthesis modelling with star formation histories derived from deep colour–magnitude diagrams constructed from Hubble Space Telescope (HST) observations. Our analysis provides a basis for understanding high-z galaxies – including those that may be unobservable even with the James Webb Space Telescope (JWST) – in the context of familiar, well-studied objects in the very low-z Universe. We find that, at the epoch of reionization, all Local Group dwarfs were less luminous than the faintest galaxies detectable in deep HST observations of blank fields. We predict that JWST will observe z ∼ 7 progenitors of galaxies similar to the Large Magellanic Cloud today; however, the HST Frontier Fields initiative may already be observing such galaxies, highlighting the power of gravitational lensing. Consensus reionization models require an extrapolation of the observed blank-field luminosity function (LF) at z ≈ 7 by at least 2 orders of magnitude in order to maintain reionization. This scenario requires the progenitors of the Fornax and Sagittarius dwarf spheroidal galaxies to be contributors to the ionizing background at z ∼ 7. Combined with numerical simulations, our results argue for a break in the UV LF from a faint-end slope of α ∼ −2 at MUV ≲ −13 to α ∼ −1.2 at lower luminosities. Applied to photometric samples at lower redshifts, our analysis suggests that HST observations in lensing fields at z ∼ 2 are capable of probing galaxies with luminosities comparable to the expected progenitor of Fornax.

Implications of Including AGB Dust in SPS Models

AbstractThe IR emission from galaxies is a unique window into multiple aspects of galaxy evolution including star-formation rates, the age of galaxies, and galactic-scale dust processes. However, asymptotic giant branch (AGB) stars continue to introduce uncertainty into stellar population synthesis (SPS) models and limit our ability to interpret the IR light of galaxies. Here we focus on incorporating circumstellar dust around AGB stars in SPS models and understanding the extent to which they influence the IR light of galaxies. We find that the significance of the AGB dust contribution depends on the characteristics of the galaxy. For quiescent galaxies and metal-poor star forming galaxies, circumstellar dust emission can have a large effect, whereas for dusty star-forming galaxies the circumstellar emission is dwarfed by emission from dust in the ISM. The models with circumstellar dust also suggest, in agreement with previous work, that IR colors can be a powerful age diagnostic for older stellar systems. Models such as these will be essential for interpreting data that will be provided by JWST and other next generation IR facilities.

Imaging of NGC 5907's stellar stream

AbstractWe have obtained deep g, r, and i-band Subaru and ultra-deep 3.6 μm IRAC images of parts of the multiply-wrapped stellar stream around the nearby edge-on galaxy NGC 5907. We have fitted the surface brightness measurements of the stream with FSPS stellar population synthesis models to derive the metallicity and age of the brightest parts of the stream. The resulting relatively high metallicity ([Fe/H] = −0.3) is consistent with a major merger scenario but a satellite accretion event cannot be ruled out.

On the uncertainties of stellar mass estimates via colour measurements

Mass-to-light versus colour relations (MLCRs), derived from stellar population synthesis models, are widely used to estimate galaxy stellar masses (M*), yet a detailed investigation of their inherent biases and limitations is still lacking. We quantify several potential sources of uncertainty, using optical and near-infrared (NIR) photometry for a representative sample of nearby galaxies from the Virgo cluster. Our method for combining multiband photometry with MLCRs yields robust stellar masses, while errors in M* decrease as more bands are simultaneously considered. The prior assumptions in one's stellar population modelling dominate the error budget, creating a colour-dependent bias of up to 0.6 dex if NIR fluxes are used (0.3 dex otherwise). This matches the systematic errors associated with the method of spectral energy distribution (SED) fitting, indicating that MLCRs do not suffer from much additional bias. Moreover, MLCRs and SED fitting yield similar degrees of random error (∼0.1–0.14 dex) when applied to mock galaxies and, on average, equivalent masses for real galaxies with M* ∼ 108–11 M⊙. The use of integrated photometry introduces additional uncertainty in M* measurements, at the level of 0.05–0.07 dex. We argue that using MLCRs, instead of time-consuming SED fits, is justified in cases with complex model parameter spaces (involving, for instance, multiparameter star formation histories) and/or for large data sets. Spatially resolved methods for measuring M* should be applied for small sample sizes and/or when accuracies less than 0.1 dex are required. An appendix provides our MLCR transformations for 10 colour permutations of the grizH filter set.

Study of the stellar population properties in the discs of ten spiral galaxies

We investigated the properties of the stellar populations in the discs of a sample of ten spiral galaxies. Our analysis focused on the galaxy region where the disc contributes more than 95 per cent of total surface brightness in order to minimise the contamination of the bulge and bar. The luminosity-weighted age and metallicity were obtained by fitting the galaxy spectra with a linear combination of stellar population synthesis models, while the total overabundance of {\alpha}-elements over iron was derived by measuring the line-strength indices. Most of the sample discs display a bimodal age distribution and they are characterised by a total [{\alpha}/Fe] enhancement ranging from solar and supersolar. We interpreted the age bimodality as due to the simultaneous presence of both a young (Age$\,\leq\,4$ Gyr) and an old (Age$\,>\,$4 Gyr) stellar population. The old stellar component usually dominates the disc surface brightness and its light contribution is almost constant within the observed radial range. For this reason, no age gradient is observed in half of the sample galaxies. The old component is slightly more metal poor than the young one. The metallicity gradient is negative and slightly positive in the old and young components, respectively. These results are in agreement with an inside-out scenario of disc formation and suggest a reduced impact of the radial migration on the stellar populations of the disc. The young component could be the result of a second burst of star formation in gas captured from the environment.

A new methodology to test galaxy formation models using the dependence of clustering on stellar mass

We present predictions for the two-point correlation function of galaxy clustering as a function of stellar mass, computed using two new versions of the GALFORM semi-analytic galaxy formation model. These models make use of a high resolution, large volume N-body simulation, set in the WMAP7 cosmology. One model uses a universal stellar initial mass function (IMF), while the other assumes different IMFs for quiescent star formation and bursts. Particular consideration is given to how the assumptions required to estimate the stellar masses of observed galaxies (such as the choice of IMF, stellar population synthesis model and dust extinction) influence the perceived dependence of galaxy clustering on stellar mass. Broad-band spectral energy distribution fitting is carried out to estimate stellar masses for the model galaxies in the same manner as in observational studies. We show clear differences between the clustering signals computed using the true and estimated model stellar masses. As such, we highlight the importance of applying our methodology to compare theoretical models to observations. We introduce an alternative scheme for the calculation of the merger timescales for satellite galaxies in GALFORM, which takes into account the dark matter subhalo information from the simulation. This reduces the amplitude of small-scale clustering. The new merger scheme offers improved or similar agreement with observational clustering measurements, over the redshift range 0 < z < 0.7. We find reasonable agreement with clustering measurements from GAMA, but find larger discrepancies for some stellar mass ranges and separation scales with respect to measurements from SDSS and VIPERS, depending on the GALFORM model used.

QUIESCENT COMPACT GALAXIES AT INTERMEDIATE REDSHIFT IN THE COSMOS FIELD. II. THE FUNDAMENTAL PLANE OF MASSIVE GALAXIES

We examine the relation between surface brightness, velocity dispersion, and size?the fundamental plane (FP)?for quiescent galaxies at intermediate redshifts in the COSMOS field. The COSMOS sample consists of ?150 massive quiescent galaxies with an average velocity dispersion of?? ? 250 km s?1 and redshifts between 0.2 < z < 0.8. More than half of the galaxies in the sample are compact. The COSMOS galaxies exhibit a tight relation (?0.1 dex scatter) between surface brightness, velocity dispersion, and size. At a fixed combination of velocity dispersion and size, the COSMOS galaxies are brighter than galaxies in the local universe. These surface brightness offsets are correlated with the rest-frame g ? z color and Dn4000 index; bluer galaxies and those with smaller Dn4000 indices have larger offsets. Stellar population synthesis models indicate that the massive COSMOS galaxies are younger and therefore brighter than similarly massive quiescent galaxies in the local universe. Passive evolution alone brings the massive compact quiescent (MCQ) COSMOS galaxies onto the local FP at z = 0. Therefore, evolution in size or velocity dispersion for MCQ galaxies since z ? 1 is constrained by the small scatter observed in the FP. We conclude that MCQ galaxies at z ? 1 are not a special class of objects but rather the tail of the mass and size distribution of the normal quiescent galaxy population.

CIRCUMSTELLAR DUST AROUND AGB STARS AND IMPLICATIONS FOR INFRARED EMISSION FROM GALAXIES

Stellar population synthesis (SPS) models are used to infer many galactic properties including star formation histories, metallicities, and stellar and dust masses. However, most SPS models neglect the effect of circumstellar dust shells around evolved stars and it is unclear to what extent they impact the analysis of SEDs. To overcome this shortcoming we have created a new set of circumstellar dust models, using the radiative transfer code DUSTY Ivezic et al. 1999, for asymptotic giant branch (AGB) stars and incorporated them into the Flexible Stellar Population Synthesis code. The circumstellar dust models provide a good fit to individual AGB stars as well as the IR color-magnitude diagrams of the Large and Small Magellanic Clouds. IR luminosity functions from the Large and Small Magellanic Clouds are not well-fit by the 2008 Padova isochrones when coupled to our circumstellar dust models, and so we adjusted the lifetimes of AGB stars in the models to provide a match to the data. We show, in agreement with previous work, that circumstellar dust from AGB stars can make a significant contribution to the IR ($\gtrsim4\mu m$) emission from galaxies that contain relatively little diffuse dust, including low-metallicity and/or non-star forming galaxies. Our models provide a good fit to the mid-IR spectra of early-type galaxies. Circumstellar dust around AGB stars appears to have a small effect on the IR SEDs of metal-rich star-forming galaxies (i.e., when A$_{\rm V}$ $\gtrsim$~0.1). Stellar population models that include circumstellar dust will be needed to accurately interpret data from the James Webb Space Telescope (JWST) and other IR facilities.

ULTRAVIOLET ISM DIAGNOSTICS FOR STAR-FORMING GALAXIES. I. TRACERS OF METALLICITY AND EXTINCTION

We have observed a sample of 14 nearby ($z \sim 0.03$) star-forming blue compact galaxies in the rest-frame far-UV ($\sim1150-2200 \AA$) using the Cosmic Origins Spectrograph on the Hubble Space Telescope. We have also generated a grid of stellar population synthesis models using the Starburst99 evolutionary synthesis code, allowing us to compare observations and theoretical predictions for the SiIV_1400 and CIV_1550 UV indices; both are comprised of a blend of stellar wind and interstellar lines and have been proposed as metallicity diagnostics in the UV. Our models and observations both demonstrate that there is a positive linear correlation with metallicity for both indices, and we find generally good agreement between our observations and the predictions of the Starburst99 models. By combining the rest-frame UV observations with pre-existing rest-frame optical spectrophotometry of our blue compact galaxy sample, we also directly compare the predictions of metallicity and extinction diagnostics across both wavelength regimes. This comparison reveals a correlation between the UV absorption and optical strong-line diagnostics, offering the first means of directly comparing ISM properties determined across different rest-frame regimes. Finally, using our Starburst99 model grid we determine theoretical values for the short-wavelength UV continuum slope, $\beta_{18}$, that can be used for determining extinction in rest-frame UV spectra of star-forming galaxies. We consider the implications of these results and discuss future work aimed at parameterizing these and other environmental diagnostics in the UV as well as the development of robust comparisons between ISM diagnostics across a broad wavelength baseline.

Star-formation histories of local luminous infrared galaxies

We present the analysis of the integrated spectral energy distribution (SED) from the ultraviolet (UV) to the far-infrared and H$\alpha$ of a sample of 29 local systems and individual galaxies with infrared (IR) luminosities between 10^11 Lsun and 10^11.8 Lsun. We have combined new narrow-band H$\alpha$+[NII] and broad-band g, r optical imaging taken with the Nordic Optical Telescope (NOT), with archival GALEX, 2MASS, Spitzer, and Herschel data. The SEDs (photometry and integrated H$\alpha$ flux) have been fitted with a modified version of the MAGPHYS code using stellar population synthesis models for the UV-near-IR range and thermal emission models for the IR emission taking into account the energy balance between the absorbed and re-emitted radiation. From the SED fits we derive the star-formation histories (SFH) of these galaxies. For nearly half of them the star-formation rate appears to be approximately constant during the last few Gyrs. In the other half, the current star-formation rate seems to be enhanced by a factor of 3-20 with respect to that occured ~1 Gyr ago. Objects with constant SFH tend to be more massive than starbursts and they are compatible with the expected properties of a main-sequence (M-S) galaxy. Likewise, the derived SFHs show that all our objects were M-S galaxies ~1 Gyr ago with stellar masses between 10^10.1 and 10^11.5 Msun. We also derived from our fits the average extinction (A_v=0.6-3 mag) and the polycyclic aromatic hydrocarbons (PAH) luminosity to L(IR) ratio (0.03-0.16). We combined the A_v with the total IR and H$\alpha$ luminosities into a diagram which can be used to identify objects with rapidly changing (increasing or decreasing) SFR during the last 100 Myr.

SIMULTANEOUS ESTIMATION OF PHOTOMETRIC REDSHIFTS AND SED PARAMETERS: IMPROVED TECHNIQUES AND A REALISTIC ERROR BUDGET

We seek to improve the accuracy of joint galaxy photometric redshift estimation and spectral energy distribution (SED) fitting. By simulating different sources of uncorrected systematic errors, we demonstrate that if the uncertainties on the photometric redshifts are estimated correctly, so are those on the other SED fitting parameters, such as stellar mass, stellar age, and dust reddening. Furthermore, we find that if the redshift uncertainties are over(under)-estimated, the uncertainties in SED parameters tend to be over(under)-estimated by similar amounts. These results hold even in the presence of severe systematics and provide, for the first time, a mechanism to validate the uncertainties on these parameters via comparison with spectroscopic redshifts. We propose a new technique (annealing) to re-calibrate the joint uncertainties in the photo-z and SED fitting parameters without compromising the performance of the SED fitting + photo-z estimation. This procedure provides a consistent estimation of the multidimensional probability distribution function in SED fitting + z parameter space, including all correlations. While the performance of joint SED fitting and photo-z estimation might be hindered by template incompleteness, we demonstrate that the latter is "flagged" by a large fraction of outliers in redshift, and that significant improvements can be achieved by using flexible stellar populations synthesis models and more realistic star formation histories. In all cases, we find that the median stellar age is better recovered than the time elapsed from the onset of star formation [abridged].