The Timescales of Star Cluster Emergence: The Case of NGC 4449

Abstract We survey the young star cluster population in the dwarf galaxy NGC 4449 with the goal of investigating how stellar feedback may depend on the clusters’ properties. Using ultraviolet (UV)–optical–near-infrared photometry obtained with the Hubble Space Telescope, we have recovered 99 compact sources exhibiting emission in the Pa β hydrogen recombination line. Our analysis reveals these sources possess masses of 10 2 < M ⊙ < 10 5 , ages of 1–20 Myr, and a color excess E ( B − V ) in the range 0–1.4. After selecting clusters with masses above 3000 M ⊙ to mitigate stochastic sampling of the stellar initial mass function, we find that our IR-selected clusters have a median mass ∼ 7 × 10 3 M ⊙ and remain embedded in their surrounding gas and dust for 5–6 Myr. In contrast, line-emitting sources selected from existing UV/optical catalogs have a median mass ∼ 3.5 × 10 4 M ⊙ and have cleared their surroundings by 4 Myr. We further find that the environment in NGC 4449 has too low pressure to drive these differences. We interpret these findings as evidence that the clearing timescale from presupernova and supernova feedback is cluster mass dependent. Even in clusters with masses ∼ 7000 M ⊙ , stochastic sampling of the upper end of the stellar initial mass function is present, randomly decreasing the number of massive stars available to inject energy and momentum into the surrounding medium. This effect may increase the clearing timescales in these clusters by decreasing the effectiveness of both presupernova and supernova feedback; neither models nor observations have so far explored such dependence explicitly. Future studies and observations with, e.g., the JWST, will fill this gap.