Context
Reconstructing the star formation history (SFH) of galaxies is critical to understanding galaxy evolution because it provides insight into the growth of stellar mass (e.g. Panter et al. 2007, Cid Fernandes 2009). Currently, the most accurate way to determine the star formation histories (SFHs) of galaxies is to use color magnitude diagrams (CMDs) in conjunction with stellar population models. Unfortunately, most galaxies are too distant to discern their stellar population from individual stars. Instead, their properties are inferred from integrated spectra, which in turn, are used to reconstruct the SFHs and to infer fundamental properties (e.g. stellar mass, star formation rates, dust extinction, metallicities, etc…). While this is the most widely used method, applied both to the local universe and the most distant galaxies, it is actually a very complicated technique, whose usefulness rests on a significant number of assumptions, not all of which have been fully calibrated. It requires stellar synthesis population models, inversion algorithms, and a realistic description of how the light from the different components (stars of various ages, gas, dust, etc) of the galaxy are mixed. Even for systems in the local Universe, this information is often scarce or absent. The primary goal of this project is to calibrate this ubiquitous technique for the first time by comparing the properties derived from CMDs of the closest galaxies with integrated spectra of the same systems. This will allow us, for the first time, to accurately assess the systematics and uncertainties of this technique.
Recently, it has become clear that the CMD method is far more robust than previously thought and is capable of providing an accurate assessment of the properties of stellar systems. This was demonstrated by the Coimbra experiment (Skillman & Gallart 2002), which compared several different methods by using CMDs in conjunction with stellar population models. Despite uncertainties in photometry from crowded fields, the degeneracy between age and metallicity (e.g. Chen 2010, Conroy 2009a,b,c, Walcher 2010), all methods converged to the same results. In turn, CMDs have been used to calibrate the integrated spectra of star clusters (Koleva et al. 2010, Cid Fernandes et al. 2010) and the bar of Large Magellenic Cloud (Alloin et al. 2002 and Fritze et al. 2007). They concluded that integrated methods were limited in extracting the age and mass fractions of stellar populations with t > 4 Gyr. Unfortunately, this is where such calibrations have stopped. Here, we propose to extend the calibrations of integrated spectra via CMDs to the nearest galaxies.
Immediate aims
We have obtained the integrated spectra of nearby galaxies for which deep, high-resolution HST imaging has already been obtained and is publicly available. These observations will give a first answer to these critical questions:
1) How does the description of the inverse problem - e.g parametrization of the star formation history, dust treatment - affect the derivation of the galaxy properties from integrated spectra?
2) Compared to CMDs what are the oldest ages and the time resolution of burst that can be detected?
3) What are the effects of sampling different region sizes, inhomogeneous stellar distributions, gas and dust on the derived star formation histories?
Sample selection
The main sample is composed by all nearby galaxies with μB<25.0 and which are close enough (D < 8Mpc) so that HST has been able to resolve the stellar populations to at least 1.5 magnitudes below the tip of the red giant branch. This restricts the population of potential targets to a sample of 38 dwarves - 4 transition dwarves, 2 spheroidal, 4 dwarf spirals, 10 irregulars and 18 starbursts - selected from the samples of 60 dwarves from the local volume of Weisz et al. 2011 and the sample of 18 starburst dwarves from Mc Quinn et al. 2010).
Weisz et al. 2011