2000ApJ...532..221S

We have mapped the barred spiral NGC 5383 using the Berkeley-Illinois-Maryland Association millimeter-wave array for observations of CO (J=1-0), the Palomar 1.5 m telescope for Hα and optical broadband, and the Kitt Peak 1.3 m telescope for near-IR broadband. We compare the observed central gas and dust morphology to the predictions of recent hydrodynamic simulations calculated using the Piner, Stone, and Teuben code. In the nuclear region, our observations reveal three peaks lying along an S-shaped gas and dust distribution: two of these are at the inner end of offset bar dust lanes at the presumed location of the inner Lindblad resonance (ILR), and the other lies closer to the nucleus. In contrast, the model predicts a circumnuclear ring, not the observed S-shaped distribution; moreover, the predicted surface density contrast between the central gas accumulation and the bar dust lanes is an order of magnitude larger than observed.

These discrepancies remain for all our simulations which produce offset bar dust lanes and indicate that the model is missing an essential process or component. A small nuclear bar might account for the discrepancy, but we rule this out using a Hubble Space Telescope NICMOS (near-IR camera and multiobject spectrometer) image: this reveals a nuclear trailing spiral, not a bar; we show that coarser resolution (i.e., ground-based images) can produce artifacts that resemble bars or rings. We conclude that the discrepancies in morphology and contrast are due to the omission of star formation from the model; this is supported by the observed high rate of central star formation (7 M_☉.yr^–1), a rate that can consume most of the accumulating gas.

As is common in similar bars, the star formation rate in the bar between the bar ends and the central region is low (0.5 M_☉.yr^–1), despite the high gas column density in the bar dust lanes; this is generally attributed to shear and shocks. We note a tendency for the H II regions to be associated with the spurs feeding the main bar dust lanes, but these are located on the leading side of the bar. We propose that stars form in the spurs, which provide a high column density but low shear environment. H II regions can therefore be found even on the leading side of the bar because the ionizing stars pass ballistically through the dust lane.