Star cluster formation and evolution in nearby starburst galaxies — I. Systematic uncertainties

The large majority of extragalactic star cluster studies done to date have essentially used two or three-passband aperture photometry, combined with theoretical stellar population synthesis models, to obtain age, mass and extinction estimates, and sometimes also metallicities. The accuracy to which this can be done depends on the choice of (broad-band) passband combination and, crucially, also on the actual wavelengths and the wavelength range covered by the observations. Understanding the inherent systematic uncertainties (the main aim of this paper) is of the utmost importance for a well-balanced interpretation of the properties of extragalactic star cluster systems. We simultaneously obtain ages, metallicities and extinction values for ∼ 300 clusters in the nearby starburst galaxy NGC 3310, based on archival Hubble Space Telescope observations from the ultraviolet (UV) to the near-infrared (NIR). We show that, for ages 6 . log(age/yr) . 9, and if one can only obtain partial coverage of the spectral energy distribution (SED), an optical passband combination of at least four filters including both blue and red passbands results in the most representative age distribution, as compared to the better constrained ages obtained from the full UV–NIR SED coverage. We find that while blue-selected passband combinations lead to age distributions that are slightly biased towards younger ages due to the well-known age–metallicity degeneracy, red-dominated passband combinations should be avoided. NGC 3310 underwent a (possibly extended) global burst of cluster formation ∼ 3×10 7 yr ago. This coincides closely with the last tidal interaction or merger with a low-metallicity galaxy that likely induced the formation of the large fraction of clusters with (significantly) subsolar metallicities. The logarithmic slope of the Vband cluster luminosity function, for clusters in the range 17.7 . F606W . 20.2