The Second Annual Lowell Observatory Fall Workshop

Solar-Type Stars:
Basic Information on Their Classification and Characterization

David R. Soderblom & Jeremy R. King

Space Telescope Science Institute
3700 San Martin Drive, Baltimore MD 21218
drs@stsci.edu

Abstract

There are three classes of stars like the Sun that are defined by Cayrel de Strobel (1996): Solar-type (or Solarlike) Stars, Solar Analogous Stars, and Solar Twins. The defining physical properties of any star are its mass, composition, and age, yet none of these is a fully or directly observable quantity. Also, angular momentum and companionship are important influences on the behavior of stars like the Sun and cannot be ignored.
We will review some of the observational limits on these fundamental properties and discuss how recent observations allow samples of solar-type stars to be defined and tabulated. Some of the questions raised are:

What is a solar analog? What is a solar twin? Who cares?
The Sun as a star: how would we judge it from afar?
How well can we measure basic stellar properties relative to solar?
Are the spectroscopic micro-differences we see significant, or just variations within a common envelope?
Is angular momentum a big deal or small potatoes?
Do we understand the basic physics of stellar evolution?

1. Solar Analogs and Twins

1.1. The Definitions

Cayrel de Strobel (1996) has recently provided definitions for stars that are similar to the Sun. These definitions are necessarily vague to some degree because they depend at least as much on how well we can measure properties as on how important that property is intrinsically. For example, a star's mass is surely its single most important datum, yet the mass can only be measured directly for stars in binaries, and for such stars we suspect the propinquity of a stellar companion influences some of what we see.

Here we specify some working definitions for these categories, based on the practical degree of observational accuracy.

1.1.1. Solar-like Stars (aka "Solar-Type''):

Solar-type stars are the least restricted category here; they are similar to the Sun in mass and evolutionary state. This means that physically they have broadly similar structure, and the presence of a convective envelope is especially important, but without totally dominant convection, as is found in M dwarfs, for example. Convection, and its interaction with rotation, is critical for the occurrence of phenomena we associate with the Sun, especially magnetically-related activity.

In practical terms, Cayrel de Strobel (1996) (and Neckel 1986a, b, c) define solar-like as 0.48 to 0.80 in B-V. Our own working definition is about F8V to K2V, or 0.50 to 1.00 in B-V. Such stars are reasonably bright, hence easy to study, and can be identified readily. They appear to share the same physics and phenomenology (see, e.g., Soderblom 1985).

1.1.2. Solar Analogs:

Solar analogs are that subset of the solar-type stars known to have detailed properties "similar'' to the Sun's. To some extent this category is meant to include stars that have the potential of being solar twins, but we do not yet know enough about them. This is probably the most difficult category to quantify, and the following are somewhat arbitrary, but are based on our present ability to infer fundamental properties of stars (see Bedding et al. [1997] for reviews of these topics):

Temperature within ~500 K of the Sun's. This is based on first-order temperatures from photometry being good to roughly 100 K (Saxner & Hammerback 1985).
Metallicity within a factor of two of solar. Again, rough metallicity values from, e.g., Strömgren photometry, are about this good (Böhm-Vitense 1981; Doyle & Butler 1990).
Within 1 magnitude of the Zero-Age Main Sequence (ZAMS). This means almost any age is permitted as long as the star is still burning hydrogen.
Any kinematics. The relationship between Galactic kinematics and age is statistical at best, so the kinematic properties of individual stars tell us nothing.
No close companion. This requirement is imposed because close companions (when the orbital period is about 10 days or less) can force the rotation to be synchronous with the orbit because of tidal effects (Mathieu & Mazeh 1988). This makes the star spin much faster than it would if it were single, and observationally the star appears extremely active no matter what the age. This requirement also avoids any issue of mass transfer within a system.

1.1.3. Solar Twins:

Solar twins should be indistinguishable from the Sun as near as we can tell. This means:

Temperature within ~10 K. Such precision can now be achieved (see below).
Metallicity within ~0.05 dex. Again, this is now possible if observations and analysis of high quality are applied (see paper by Nissen in Bedding et al. 1997).
Age within ~1 Gyr, so that the evolutionary state is comparable.
No known stellar companion, since the Sun has none.

1.2. Solar Twins: Who Cares?

Does it really matter if we can't find two stars that appear to be identical? We don't expect to find a twin to ourselves, do we?

This use of the word "twin'' brings to mind human twins, and perhaps that analogy is misleading. For example, consider some twins I know, the Brown brothers, Jeremy and Zach. Both are Boy Scouts. How do I tell them apart when they're both in uniform?

For a time one wore glasses, and the other didn't, but now both now wear contacts.
For a time, one was Star Scout and the other Life, now both are Life Scouts.
Freckles, maybe? Zits?

The answer, of course, is there are always differences if one looks closely enough, yet Jeremy and Zach unambiguously fulfill any reasonable definition of identical twins. In the case of biological twins, we have a clear notion of twinness based on having identical DNA. Stars are different, of course, and not just because we're dealing with physics instead of biology. In seeking solar twins we are not seeking a common origin, but instead the same outcome, which is a very different thing after billions of years have elapsed, and considering that the stars we are comparing are certain to have been born in very different places. This leads to:

Axiom: All stars are different. If you can't tell them apart, you just haven't looked closely enough yet.

Question: What does it mean for stars to be twins? How much may they differ and still be twins for all practical purposes?

Postulate: "The same'' means that stars may vary, but they do so within a common envelope. For example, the Sun is still the Sun when it's in a Maunder minimum state or as it goes through its activity cycle, and so stars may differ in some details yet remain within a family of characteristics. The Sun defines the envelope.

Flawed reasoning 1: Solar "twins'' are preferred targets for planet searches. We do not yet know enough to rule out any class of stars as parents of planets. (But there are valid observational reasons for preferring solar-type stars because they have narrow lines and are photometrically quiet; see below.)

Flawed reasoning 2: We can find a star that we can use as a spectroscopic or spectrophotometric stand-in for the Sun, without reservations or doubt. How can we ever know for sure the proxy is identical?

[Next]
[Contents]

Solar-Type Stars: Basic Information on Their Classification and Characterization