Dating Stars as Models
Many have dreamed of dating a star, but the way astronomers do it is less glamorous. For one thing, they need to know how old she is first, and how good a model she makes. In a Perspectives piece for Science,1 David R. Soderblom of the Space Telescope Science Institute explained the requirements for stellar dating in an article entitled, “How Old Is That Star?”
It’s not that simple lining up a date. Many agents tend to get in the way:
Determining how long it has been since a star formed is a lot harder than it seems like it ought to be, and many very basic questions hinge on stellar ages. For instance, we’d like to know the ages of stars that have planets. We hope to detect signs of life on planets around other stars, but if we do, knowing the star’s age is central to interpreting what is observed. Among the youngest stars, we’d like to know how long it takes for planetary systems to form and evolve. On a grander scale, the ages of stars and clusters of stars are needed to infer the history of our Milky Way galaxy and the pieces from which it was built. Did the halo of our galaxy form on its own or from fragments of captured satellite galaxies? Did the thick disk form after the halo or contemporaneously? Has the thin disk (of which the Sun is a member) formed stars continuously, or in episodes? Many very basic questions can only be addressed if we can establish reliable ages.
The parameters we measure are in the present (with regard to the time of light travel from source to observer): mass, chemical composition, magnetic field, and other characteristics teased from the spectrum. Soderblom claims “we can measure a precise and exact age for just one star–the Sun–and that’s because we can analyze solar system material in the laboratory, something we can do for no other star.” Even that age, however, is highly model dependent. That makes estimates of other stars’ ages derived from it even further model dependent:
By calibrating models against the Sun, we can comprehend stars that are both more and less massive. Our understanding of the evolution of stars is closely tied to studying star clusters, groups of hundreds to thousands of stars that were formed together and so share the same composition and age. Or do they? Some of the most exciting astrophysics from the Hubble Space Telescope has been the discovery of multiple populations within single globular clusters, which are some of the oldest components of the Milky Way. Given what we now know about stellar physics, the available explanations include multiple ages (i.e., several epochs of star formation spread well apart in time), very different compositions of the cluster’s members, or both. Neither alternative satisfactorily explains the observations, and a very basic conundrum has been exposed.
The model uncertainties are on the order of 10-20%, he claimed, but one should also keep in mind that these uncertainties have “poorly understood systematic effects.” Relative ages are more believable, he said. Even when using radiometric ages, the derived dates must be interpreted from when the isotopes formed without knowing the initial abundances. These do not necessarily reflect the age of the star.
Soderblom examines some of the “empirical” measurements for dating stars: loss of angular momentum over time, and asteroseismology (oscillations). Both these methods are also model dependent: “We can see a consistent relation between a physical quantity and age, but we do not understand the underlying physical relation, even though we may have at least a reasonable scenario.” What is considered reasonable becomes subjective. Some techniques seem more “promising” – language indicative that the key component of the spectrum may be the human element.
“Overall, the situation for determining stellar ages is still sobering, and progress has been slow,” he ended. It has reached the point where cosmologists claim better precision for their measurements than we can for the ages of the nearest and brightest stars.” He did not distinguish between claims and realities, however, since cosmological claims are also highly model dependent. Within his own subject matter, “The challenge of determining an accurate age for a star therefore remains outstanding.”
1. David R. Soderblom, “Astronomy: How Old Is That Star?”, Science, 2 January 2009: Vol. 323. no. 5910, pp. 45-46, DOI: 10.1126/science.1168230.
People need to know the hand-waving and speculation that belies the confidence expressed on TV science specials. One can detect enough wiggle room to permit major paradigm shifts. Notice that model dependence is not observation dependence. We can observe emanations from objects that hit our eyeballs in the present; what does that mean about their origins and histories? One cannot know that without making assumptions. The reasonableness of assumptions is a matter of opinion. It may seem reasonable to you, but if a member of the Space Telescope Science Institute says that “The challenge of determining an accurate age for a star therefore remains outstanding,” that should at least be noted in the minutes.