Initial Mass Function

The IMF $\xi(M)$ specifies the distribution in mass of a newly formed stellar population and it is frequently assumed to be a simple power law: $\xi(M) = c \, M^{-(1+x)}$. In general, $\xi(M)$ is assumed to extend from a lower to an upper cutoff, chosen to be $M_1=0.1\,M_\odot$ and $M_2=125\,M_\odot$ in GISSEL. In Table 1 we show the parameters of the three most used IMFs: the Salpeter (1955), Scalo (1986) and Miller & Scalo (1979) laws, in the form used by Bruzual & Charlot in their evolutionary synthesis model.
Table:Parameters of Salpeter (1955), Scalo (1986) and Miller & Scalo (1979) laws for the IMF: $M_1$ and $M_2$ are the lower and higher mass cutoffs, and $x$ is the parameter of the power law. 
IMF $M_1$ $M_2$ $x\quad$
Salpeter 0.10 125. 1.35
Scalo 0.10 0.18 -2.60
  0.18 0.42 0.01
  0.42 0.62 1.75
  0.62 1.18 1.08
  1.18 3.50 2.50
  3.50 125. 1.63
Miller & Scalo 0.10 1.00 0.25
  1.00 2.00 1.00
  2.00 10.0 1.30
  10.0 125. 2.30
\begin{figure}\centerline {\psfig{,width=0.92\textwidth,angle=270}}\end{figure}
Figure:The three IMFs used in the spectral evolutionary models of Bruzual & Charlot (1993).
The different slopes of the three considered laws produce different spectral energy distributions: the Scalo and Miller & Scalo laws are flat at small masses and less rich of massive stars with respect to the Salpeter law, as illustrated in Figure 2. The large number of massive stars in the Salpeter law produces an excess of UV flux, whereas the Scalo law generates too many solar mass stars, making the spectrum too red to match the observed colours.

We adopted the Miller & Scalo law as a good compromise. We also tested the influence of a change in IMF, finding a negligible effect in photometric redshift computation (Bolzonella et al. 2000).

micol bolzonella