Genevieve J. Graves: Research  
  Truncated Star Formation and the 2D Family of Early Type Galaxy Star Formation Histories hyperplane_image

Papers I and II in this series demonstrated that the stellar populations of quiescent, non-star-forming galaxies populate a 2D parameter space. Here, we quantify the mapping of these stellar population variations onto a cross-section projected through the Fundamental Plane (FP). The first and well-known dimension of variation correlates with velocity dispersion, such that higher-velocity dispersion galaxies have older luminosity-weighted mean ages, higher [Fe/H], higher [Mg/H], and higher [Mg/Fe] than those with lower velocity dispersions. The second dimension maps onto a cross-section through the FP at fixed velocity dispersion and fixed effective radius (R), such that galaxies that scatter off the FP to higher surface brightnesses (higher Delta I) are younger and more metal-rich in both [Fe/H] and [Mg/H], but are less Mg-enhanced than their counterparts on the FP midplane. The 2D family of stellar population variations can be interpreted as a 2D family of star formation histories in which both the start time and duration of star formation depend on the velocity dispersion, but where the duration of star formation varies among galaxies of the same velocity dispersion. At fixed velocity dispersion, galaxies with longer duration star formation have higher stellar surface mass densities, higher metallicities ([Fe/H] and [Mg/H]), and higher total stellar mass for a given size. This suggests that these galaxies have more thoroughly processed gas into stars. The correlation between higher ``conversion efficiency'' of gas into stars and longer durations for star formation strongly supports the ``premature truncation'' scenario for galaxy formation we proposed in Paper III. In this model, galaxies of a given velocity dispersion experience similar star formation histories, up until they have their star formation truncated at different times. Those truncated earlier are shut down while their formation is incomplete. As a result, they form fewer stars, have lower stellar mass densities, and have lower metal abundances. We show that this model is quantitatively consistent with simple expectations for chemical enrichment in galaxies. For more information, keep an eye on astro-ph!

  Mass-to-Light Ratio Variations in 3D Fundamental Plane Space galaxy_image

The Fundamental Plane of early type galaxies is observed to have finite thickness and to be tilted from the virial relation. Both of these represent departures from the simple assumption that dynamical mass-to-light ratios (Mdyn/L) are constant for all early type galaxies. We use a sample of 16,000 quiescent galaxies from the Sloan Digital Sky Survey to map out the variations in Mdyn/L throughout the 3D Fundamental Plane space defined by velocity dispersion, effective radius (R), and effective surface brightness (I). Dividing Mdyn/L into multiple components allows us to separately consider the contribution to the observed Mdyn/L variation due to stellar population effects, IMF variations, and variations in the dark matter fraction within one R. The observed tilt of the FP requires contributions from both, with dark matter/IMF variations likely comprising the dominant contribution. The same is true through the thickness of the FP. This means that the finite thickness of the FP is due to variations in the stellar mass surface density within R, not the fading of passive stellar populations and it therefore represents genuine structural differences between early type galaxies. These structural variations are correlated with galaxy star formation histories such that galaxies with higher Mdyn/M* have higher [Mg/Fe], lower metallicities, and older mean stellar ages. We discuss several physical mechanisms that might explain the observed co-variation between Mdyn/M* and galaxy star formation histories. It is difficult to explain the observed enhancement of alpha-elements in lower-surface-brightness galaxies by allowing the IMF to vary. Differences in dark matter fraction can be produced by variations in the ``conversion efficiency'' of baryons into stars or by the redistribution of stars and dark matter through dissipational merging. The former explanation, specifically a model in which some galaxies experience low conversion efficiencies due to premature truncation of star formation, provides a more natural explanation for the co-variation of Mdyn/M* and the observed stellar population properties. For more information, keep an eye on astro-ph!

  Star Formation Histories Throughout the Fundamental Plane galaxy_image

How is the present day structure of a galaxy connected to its past star formation history? We know that early type galaxies populate a relatively tight 2-dimensional plane in the 3-dimensional parameter space described by velocity dispersion, effective radius (R), and effective surface brightness (I). This is known as the Fundamental Plane. We ask the question: do early type galaxies also populate a multi-dimensional space in terms of their star formation histories, analogous to the Fundamental Plane? To address this, we study the stellar populations of galaxies in 3-D Fundamental Plane space and track variations as a function of velocity dispersion, R, and I. We find that early type galaxies do indeed form a 2-D family in stellar population space. However, rather than this 2-D space corresponding to the face-on Fundamental Plane, it in fact corresponds to slices through the thickness of the plane: stellar populations vary with velocity dispersion (as is well known), but vary also with surface brightness for fixed values of velocity dispersion and R. Thus, although the Fundamental Plane is quite tight, outlying galaxies which fall above or below the plane (i.e., at higher or lower I) appear to have experienced different star formation histories than the galaxies which fall on the midplane of the Fundamental Plane, such that lower surface brightness galaxies have experienced shorter durations of star formation, while higher surface brightness galaxies have experienced more prolonged star formation than corresponding galaxies on the midplane. For more information, see Graves et al., 2009b, ApJ, 698, 1590.

  The Color-Magnitude vs. The Color-Sigma Relation galaxy_image

The properties of galaxies are often studied as a function of galaxy "size". Different size measurements (stellar mass, luminosity, stellar velocity dispersion) are used roughly interchangeably. However, due to correlated residuals, the stellar populations of galaxies depend in different ways on these various size measurements, leading to results between different studies that appear to be in conflict. By explicitly separating out the stellar population dependencies on velocity dispersion, luminosity (L), and color, we are able to show that velocity dispersion is a better predictor of a galaxy's past star formation history than is L. Stellar population age, metallicity, and enhancement in alpha-elements all correlate strongly with galaxy velocity dispersion. The residual variations in stellar population age and metallicity at fixed velocity dispersion are correlated with one another and with L in such a way as to strengthen the existing metallicity trends and counter-act the existing age trends, leading to a strong L-metallicity relation and a very weak L-age relation. These correlated residuals at fixed velocity dispersion have opposing effects on the galaxy color which tend to cancel out such that there is no color-magnitude relation at fixed velocity dispersion: the observed color-magnitude relation is entirely the result of color-velocity dispersion and magnitude-velocity dispersion relations. For more information, see Graves et al., 2009a, ApJ, 693, 486.

  Stellar Population Modelling: Fe, Mg, C, N, and Ca galaxy_image

Recent improvements in the modelling of stellar atmospheres has made it possible to determine the effects of non-Solar abundance patterns on stellar absorption-line spectra. Ricardo Schiavon has taken advantage of these abundance sensitivity calculations to construct stellar population models that include the effects of non-Solar abundances for the elements C, N, O, Fe, Mg, Ca, Na, Si, Cr, and Ti. These models, described in Schiavon (2007), make it possible to dial in your favorite abundance pattern and get out predicted absorption line strengths for a model stellar population. In practice, one most often wants to do the reverse: take an observed set of line strengths and determine the abundance pattern of the input galaxy. We have implemented a method that automates this process and made it publicly available in an IDL code called "EZ_Ages". The method is presented in Graves & Schiavon 2008, ApJS, 177, 446, along with extensive tests to demonstrate that it works. To download the code and instructions for its use, click here.

  Stellar Populations in Red Sequence LINERs galaxy_image

A significant fraction of early type galaxies show emission lines characteristic of Low-Ionization Nebular Emission Regions, or LINERs. I used ~6000 spectra from the Sload Digital Sky Survey (SDSS), combining many individual spectra to achieve high signal-to-noise, in order to study the stellar populations of these LINERs, as compared to similar early type galaxies which lack emission. The LINERs have similar metallicities and element abundance ratios, but they are typically ~2 billion years younger than corresponding early type galaxies without emission. These ages represent a luminosity-weighted average of the stars in the galaxy, rather than a true age estimate, but they suggest that LINERs were forming stars up until more recently than their emission-free counterparts. If the LINER emission in these galaxies is powered by low-level accretion onto a black hole at the galaxy's center, this may indicate a link between the star formation history of a galaxy and the activity of its massive black hole. However, the LINER emission may be distributed in many of these galaxies rather than being concentrated around the galaxy center, suggesting that the emission may be powered by post-AGB stars. For more information, see Graves et al. 2007, ApJ, 671, 243.

  SN 1987A: Still no Neutron Star! supernova_image

My undergraduate senior thesis research used HST/ACS images and HST/STIS spectroscopy to put an upper limit on a point source remnant of the SN1987A progenitor. There is still no pulsar in the supernova remnant to date, and no visible point source. Our upper limit (L < 8 x 10^33 ergs/s) was tight enough that we were able to restrict possible accretion scenarios onto a compact remnant in the center of the supernova. Spherical accretion onto either a neutron star or a black hole appears to be incompatible with our limit. If an accretion disk exists in the remnant, it is limited to a low mass accretion rate (< 0.3 Eddington) at a small radius (r < 10^10 cm). For more information, see Graves et al. 2005, ApJ 629, 944.

  Home   Research   Publications   EZ_Ages   CV