The ages of pre-main-sequence stars

The position of pre-main-sequence or protostars in the Hertzsprung-Russell diagram is often used to determine their mass and age by comparison with pre-main-sequence evolution tracks. On the assumption that the stellar models are accurate, we demonstrate that, if the metallicity is known, the mass obtained is a good estimate. However, the age determination can be very misleading, because it is significantly (generally different by a factor of 2 to 5) dependent on the accretion rate and, for ages less than about 10(6) yr, the initial state of the star. We present a number of accreting protostellar tracks that can be used to determine age if the initial conditions can be determined and the underlying accretion rate has been constant in the past. Because of the balance established between the Kelvin-Helmholtz, contraction time-scale and the accretion time-scale, a pre-main-sequence star remembers its accretion history. Knowledge of the current accretion rate, together with an HR-diagram position, gives information about the rate of accretion in the past, but does not necessarily improve any age estimate. We do not claim that ages obtained by comparison with these particular accreting tracks are likely to be any more reliable than those from comparisons with non-accreting tracks. Instead, we stress the unreliability of any such comparisons, and use the disparities between various tracks to estimate the likely errors in age and mass estimates. We also show how a set of coeval accreting objects do not appear coeval when compared with non-accreting tracks. Instead, accreting pre-main-sequence stars of around a solar mass are likely to appear older than those of either smaller or larger mass.