2004A&A...421..887I

Since the Hα and UV fluxes from galaxies are sensitive to stellar populations of ages <10⁷ and ≃10⁸yr respectively, their ratio f(Hα)/f(UV) provides us with a tool to study the recent t≤10⁸yr star formation history of galaxies, an exercise that we present here applied to 98 galaxies in 4 nearby clusters (Virgo, Coma, Abell 1367 and Cancer). The observed f(Hα)/f(UV) ratio is ∼ a factor of two smaller than the expected one as determined from population synthesis models assuming a realistic delayed, exponentially declining star formation history. We discuss various mechanisms that may have affected the observed f(Hα)/f(UV) ratio and we propose that the above discrepancy arises from either the absorption of Lyman continuum photons by dust within the star formation regions or from the occurrence of star formation episodes. After splitting our sample into different subsamples according to evolutionary criteria we find that our reference sample of galaxies unaffected by the cluster environment show an average value of f(Hα)/f(UV) two times lower than the expected one. We argue that this difference must be mostly due to absorption of ≃45% of the Lyman continuum photons within star forming regions. Galaxies with clear signs of an ongoing interaction show average values of f(Hα)/f(UV) slightly higher than the reference value, as expected if those objects had SFR increased by a factor of ≃4. The accuracy of the current UV and Hα photometry is not yet sufficient to clearly disentangle the effect of interactions on the f(Hα)/f(UV) ratio, but significant observational improvements are shortly expected to result from the GALEX mission.