Dr. James Gunn (Princeton University) will undertake a detailed examination of the ingredients and methodology of stellar population synthesis (SPS) models. This type of modeling provides the bridge between the physical properties of galaxies (such as chemical composition, stellar mass, and star formation history) and the observed spectral energy distributions. It also represents the fundamental link between extragalactic observations and models of galaxy assembly. Despite its importance in many areas of extragalactic astronomy, the uncertainties associated with SPS modeling have received relatively little attention. This situation is especially troubling in light of known and important deficiencies in current implementations of SPS.
In this project, Dr. Gunn will develop a flexible SPS code that is capable of incorporating uncertainties in stellar evolution, including the evolution and properties of evolved stars of all types. He will also explore uncertainties in stellar spectral libraries, models of dust in galaxies, the distribution of stellar abundances and the history of chemical enrichment, and the relative numbers of stars of different mass. The goals of the study are to provide robust constraints on the physical properties of large samples of galaxies, to investigate the extent to which uncertainties in SPS make it difficult to compare galaxy formation models to observations, and to highlight which uncertainties in SPS are the most important to address with future observations.
The new, versatile SPS code developed under this proposal will be publicly released so that it may be maximally useful to the community. In addition, the code will be open source, which is unique for SPS codes. The project will also create comprehensive catalogs of a variety of physical properties of galaxies from existing and future surveys of galaxy properties. These catalogs will be fully released to the public on a short time-scale.
We have learned in the last few decades that ordinary matter, made of protons, neutrons, and electrons, comprises only about 4 percent of the total mass/energy of the universe; the vast majority is in two still completely mysterious forms, dark matter (23 percent), which interacts,as far as we know, only gravitationally, and dark energy, (73 percent) which is responsible for the observed acceleration of the expansion of the universe. Only about half of the ordinary matter is in galaxies like our own Milky Way, but it is these galaxies, via the stars and gas in them, that tell us almost everything we know about the present properties of the universe and how it has evolved into its present state. It is thus of great importance to understand the properties of galaxies and the evolution of those properties through cosmic time. This project was designed to deal with one aspect of this problem, the study of the populations of stars which make up galaxies, called "stellar population synthesis." The subject is an old one, beginningin the 1970s, but technological advancements in computing and detectors have made large strides possible in the last decade. What we have done in this project are small steps in the direction of laying a firmer foundation, and, in addition, finding quantitative ways to assess the very large uncertainties in the approaches based on theusual assumptions. These assumptions are necessary because of our ignorance of several vital properties of stars and galaxies, but we need to understand how serious our ignorance *is*. In order to take these uncertainties into account, first a very flexible computer code was developed which allowed the explicit recognition of these problems with attempts to set reasonable limits on how much our ignorance of them can affect the synthesis results. There have been many synthesis codes developed over the years, but none until ours had this flexibility. The code was also made public in its entirety, so other workers can modify it freely if they disagree with our methods or estimates. Second, the code was used to study nearby systems with very good data, to attempt actually to remove some of the uncertainties; this met withsome success, in particular in the study of a particularly troublesome phase in the evolution of stars slightly more massive than the sun, the so-called "thermally pulsating asymptotic giant branch" stars, which go through a very short, bright phase.. The final chapter in this study was to apply these techniques to study one aspect of the synthesis models--namely, how much mass there is in stars after star formation and evolution has gone on for the age of the universe. We do this by measuring the rotation velocity of galaxiesby looking at the differential redshift from one side to the other, to measure the orbital speeds of stars about the center of the galaxy. It was discovered years ago by Tully and Fisher that the rotation velocity of galaxies was correlated in a very tight relationship to the brightness of the galaxy. Not surprisingly, little galaxies with less mass rotate more slowly than big, massive ones. What *was* surprising was that the relation was so regular and so accurately predictive. What we have done here is thee-pronged. First, we have rederived theTully-Fisher relation with more modern instruments for a large sample of galaxies. Since it is a statistical relation, the construction of the sample is very important. We also constructed a more distant sample of very similar objects using the survey data from the Sloan Survey. Second, we have used this more distant sample to study statistically the bending of light (gravitational lensing) from yet more distant galaxies as their light passes the galaxy in our sample. This bending allows us to measure the *total* mass--stars plus dark matter--in the sample galaxy. Finally, we then appeal to computer models to determine the dark matter contribution to the masses we observe from the rotation velocities and obtain the mass in stars; these data can be compared with computer-generated cosmological models of the distribution of mass in the universe. The conclusion is tentative as yet, and waits on much larger and deeper surveys than the Sloan (which are coming), but with something like twenty percent accuracy, we can say that the stellar masses are correct and that the standard cosmological model makes predictions about dark matter structures which are basically correct. It is quite certain that our future (better) understanding of theuniverse, its structures, its basic contents, and its dynamics, will depend more and more on our detailed understanding of galaxies and the stars, gas, dust, and dark matter which make them up. This study attempts basically to increase our knowledge of the stellar contribution, and hence the source of most of the light by which we study the galaxies.
