1.3 How can we explain the distribution of stars on the H-R diagram?
Here is a possible explanation for the concentration of stars into certain regions on the H-R diagram. It is based on the reasonable assumptions that:
Any particular star is luminous for only a finite time;
There are distinct stages between the star's cradle and grave, each stage being characterized by some range of temperature and luminosity; the star thus moves around the H-R diagram as it evolves;
The stars we see today are not all at the same stage of evolution.
From these reasonable assumptions it follows that if we observe a large population of stars today, then the longer a particular stage lasts the greater will be the number of stars that are observed in that stage. Conversely, we will catch very few stars going through a short-lived stage.
We can thus explain the concentrations on the H-R diagram as those regions where the stars spend a comparatively large fraction of their lives. On this basis a star must spend most of its life on the main sequence, because this is where about 90% of the stars lie. Where it lies before it joins the main sequence, and where it goes afterwards, we cannot tell without further information, but the red giant, supergiant and white dwarf regions are where, on our assumptions, we might expect some stars to dwell for a while.
There are two other factors that influence the concentrations of stars on the H-R diagram. First, the concentration depends not only on how quickly a star passes through a region, but also on what fraction of stars pass through the region at all. Second, some regions of the H-R diagram might be bereft of stars simply because they correspond to stages in a stellar lifetime when stars tend to be shrouded in cooler material and are therefore not observable directly.
We clearly need more observational data to make further progress. Observations of individual stars actually evolving would be of enormous value. Can we see such evolution by making observations over a period of time?
Unfortunately, with very few exceptions, we can't. This is because stars evolve extremely slowly. We have good evidence that the Sun is about 4.5 × 109 years old, and that it will be about as long again before it runs out of hydrogen fuel in its core. The lifetime of an astronomer, or indeed the whole history of astronomy, are both tiny fractions of this 4.5 × 109 year timescale. Changes in the Sun and other stars in short times are usually small. No matter how obvious the changes in the Sun are to us, if we had to view the Sun as a star from a great distance they would be insignificant. However, some stars do change on short timescales - the spectacular supernovae and variable stars.
One type of supernova, the Type II supernova, marks the end of a supergiant star. Thus, Betelgeuse and Rigel A seem fated to disappear after a final blaze of glory, their luminosity rising 108 times in a few days, followed by a few months of decline into oblivion, when they will vanish from the sky and from the H-R diagram.
All types of novae, which exhibit one or more short-lived outbursts, are in binary systems. In a minority of binary systems the two stars are so close together that they interfere with each other's evolution. In some cases, this will lead to one of the two undergoing a nova outburst. Observations of novae thus help us to understand disturbances to the normal course of stellar evolution, and this also helps us to understand the normal course itself.
The irregular variable T Tauri stars lie just above the main sequence on the H-R diagram (Figure 7) in a zone that covers a wide range of temperatures, including that of the Sun, and they lie among traces of the sort of interstellar material from which stars are thought to form. These observations suggest strongly that they are very young stars, about to settle on to the main sequence. Indeed, some T Tauri stars probably have been seen to do just this. Therefore, the early phase of stellar evolution can be elucidated by the study of these stars.
Cepheids and other types of regular pulsating variable stars also help us to understand some of the processes that drive evolution at certain stages in a star's life.
Have we now exhausted the main sources of observational data that help us to build models of the stars and of their evolution? No, there is one further property of enormous importance, and this is a star's mass.
If most stars were to end their lives quietly, by gradually cooling at roughly constant radius, what sort of tracks would they make across the H-R diagram?
Such stars would move diagonally to the right and downwards, the luminosity as well as the temperature decreasing. You will see later that many stars do indeed end their lives in this way.