This paper aims to see whether the observed distribution of the Roberts time for different samples of late-type galaxies of a given class (normal, active) may be fitted by a single theoretical distribution, and what additional constraints are put on models obeying a generalized Schmidt star formation law. We find that an extended Poisson distribution with expected value k*=2.5Gyr and median value k‡=2.83Gyr is consistent with five different samples of normal galaxies, provided systems with Roberts time TR >7Gyr are excluded, and an extended Poisson distribution with expected value k*=0.05Gyr and median value k‡=0.52Gyr is consistent with both a sample of active galaxies and a sample of compact HII regions. In this view, active galaxies and ``normal'' galaxies with TR >7Gyr should be regarded as merger products in different stages of evolution. In dealing with a generalized Schmidt star formation law, we take into consideration closed, comoving models of chemical evolution. The history of a galaxy is described by two main phases: contraction (which produces the extended component) and equilibrium (which gives the disk). Negative values of the ratio of contraction time to age of a galaxy, Tc/T, relate with present-day, fractional star formation rates CD <~0.025G/yr and Roberts times TR>5Gyr but, in general, the reverse is not true. Again, an extended Poisson distribution with expected value k*=2.5Gyr and median value k‡= 2.83Gyr is consistent with two different samples (for which total masses also are reported and then ratios Tc/T can be calculated) of normal galaxies, provided systems for which Tc/T <=0 are excluded. In this view, active galaxies and ``normal'' galaxies for which Tc/T <=0 should be regarded as mergers products in different stages of evolution. Further analysis provides additional support to two conclusions already established in a previous paper, namely: (i) models where star formation obeys a pure Schmidt law cannot fit the whole set of observations; and (ii) for assumed TR=k‡; Tc/T>0 for normal galaxies; and Tc/T<0.5 for at least a fraction of large-mass (M>1011Msun) normal galaxies; a generalized Schmidt star formation law with an exponent n=~1 is favoured over one with n=~2. More precisely, 1<~n<~1.8 with a preferred value n=~1.2, which makes a generalized Schmidt star formation law with n=~1 consistent with the observations. Finally, a stochastic process of star formation, related to a Poisson distribution, is briefly outlined.
A generalized Schmidt star formation law: observational constraints. II
CAIMMI, ROBERTO
1996
Abstract
This paper aims to see whether the observed distribution of the Roberts time for different samples of late-type galaxies of a given class (normal, active) may be fitted by a single theoretical distribution, and what additional constraints are put on models obeying a generalized Schmidt star formation law. We find that an extended Poisson distribution with expected value k*=2.5Gyr and median value k‡=2.83Gyr is consistent with five different samples of normal galaxies, provided systems with Roberts time TR >7Gyr are excluded, and an extended Poisson distribution with expected value k*=0.05Gyr and median value k‡=0.52Gyr is consistent with both a sample of active galaxies and a sample of compact HII regions. In this view, active galaxies and ``normal'' galaxies with TR >7Gyr should be regarded as merger products in different stages of evolution. In dealing with a generalized Schmidt star formation law, we take into consideration closed, comoving models of chemical evolution. The history of a galaxy is described by two main phases: contraction (which produces the extended component) and equilibrium (which gives the disk). Negative values of the ratio of contraction time to age of a galaxy, Tc/T, relate with present-day, fractional star formation rates CD <~0.025G/yr and Roberts times TR>5Gyr but, in general, the reverse is not true. Again, an extended Poisson distribution with expected value k*=2.5Gyr and median value k‡= 2.83Gyr is consistent with two different samples (for which total masses also are reported and then ratios Tc/T can be calculated) of normal galaxies, provided systems for which Tc/T <=0 are excluded. In this view, active galaxies and ``normal'' galaxies for which Tc/T <=0 should be regarded as mergers products in different stages of evolution. Further analysis provides additional support to two conclusions already established in a previous paper, namely: (i) models where star formation obeys a pure Schmidt law cannot fit the whole set of observations; and (ii) for assumed TR=k‡; Tc/T>0 for normal galaxies; and Tc/T<0.5 for at least a fraction of large-mass (M>1011Msun) normal galaxies; a generalized Schmidt star formation law with an exponent n=~1 is favoured over one with n=~2. More precisely, 1<~n<~1.8 with a preferred value n=~1.2, which makes a generalized Schmidt star formation law with n=~1 consistent with the observations. Finally, a stochastic process of star formation, related to a Poisson distribution, is briefly outlined.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.