Quantifying the poor purity and completeness of morphological samples selected by galaxy colour

ABSTRACT The galaxy population is strongly bimodal in both colour and morphology, the two measures correlate strongly, with most blue galaxies being late-types (spirals) early-types, typically ellipticals, red. This observation has led to use of as a convenient selection criterion make samples that are then labelled by morphology. Such proxy for morphology results necessarily impure incomplete samples. In this paper, we morphological labels produced Galaxy Zoo measure how such are, considering optical (ugriz), near-ultraviolet (NUV), near-infrared (NIR; JHK) bands. best single found using threshold g ? r = 0.742, but still sample where only 56 per cent red smooth Use NUV gives some improvement over purely bands, particularly late-types, low purity/completeness early-types. No significant adding NIR With any including NUV, early-types greater than two-thirds purity cannot be constructed. Advances quantitative morphologies have made colour–morphology selections largely unnecessary going forward; assumptions required, recommend studies carefully consider implications incompleteness/impurity.