As I do chemical evolution models for spiral galaxies, I am computing the SED's of these kind of galaxies using directly the star formation history and the age-metallicity relation provided by the results of our code as the SSP's for calculating the SED. (The inverse technique that the usual one, I think!)

n fact in a very simple way I have done a figure (sp2_scott.eps) with the spectrum2 and our model for the most massive spiral and the highest star formation efficiency in order to check how well (or bad!) goes and I think it is not so bad...  I have not taken into account any dispersion velocity nor other calculations, except the normalization, only drawing what I have over the data.  Although, like this is a model for a spiral galaxy, I have taken the results for three radial regions located at 4, 8 and 12 kpc of the center and I have done an averaged spectrum...

I send you the file spgal_n4dis28_r8kpc for the spectra corresponding to a galaxy like MWG, Sbc, with a rotation velocity of 200 km/s . The file has 2 columns. The first one is lambda (in AA), the second one the spectrum the radial region of R=8 kpc which would correspond to the solar vicinity, for which we know the star formation history and the age-metallicity relation with quiet precise measurements....I have the spectra for more radial regions, since R=2 kpc up R=24 kpc so if you prefer another...

The file n4dis28genR8kpc gives for each time step (in Gyr), column 1, the mass of stars created in each time (19^9Msun) in column 2), the total abundance Z (column 3) and some elemental abundances in the following ones. The last one is [Mg/Fe].  With this file, the SFH and Z(t) are well defined.

Added 27 October 2006:

We have found an error in the chemical evolution model we have taken to do the figure I send you in september (which correspond to sp2 of test3) . We have repeated this figure, and we have added two other panels to show [Fe/H] (different of the total Z) and [Mg/Fe].

Figure
Spectrum (old)
Model output (old)