1. OPACITY IN THE LOCAL AND DISTANT UNIVERSE

Dust opacity alters the radiation from astronomical sources and the physical quantities we derive from it. The effects of dust on the light are twofold: a) global dimming of the radiation output from a source and b) selective (i.e., wavelength-dependent) removal of the radiation. Because of the wavelength dependence of the extinction, the UV will be more affected by dust than the optical or the infrared; sources which are optically thin in the visible (e.g., A_V = 0.3) become optically thick in the UV(A_{1300 Å} 1). Dust, therefore, can have a major effect on the wavelength range selected for studying high-mass star formation processes. Matters are further complicated by two additional "characteristics" of reddening:

• the details of the extinction curve are environment-dependent. The best known cases are the diffuse ISM extinction curves of the Milky Way [36], the Large Magellanic Cloud (LMC) [15], the Small Magellanic Cloud (SMC) [4], and our twin galaxy M31 [2], which are characterized by different shapes. Even within our own Milky Way, different environments have different extinction curves [11].

• In extended objects (HII regions, galaxies, etc.) the 'effective' obscuration is a function of the geometrical distribution of the dust relative to the emitters (e.g., [43]). An extinction curve can be defined or used in a straightforward manner only when all the dust is foreground to the emitter, as in the case of stars. In galaxies, the dust is usually mixed in a complicate fashion with the dust; geometry becomes the dominant factor in determining the wavelength dependence of the obscuration [29, 8]. Even in the simplest (unrealistic) case that all the dust is in a shell surrounding the galaxy, back-scattering from the farthest regions of the dust shell into the line of sight produces an effective obscuration which is greyer than the 'standard' extinction curves.

While it is generally agreed that galaxies, at least the late Hubble types, contain dust, less agreement exists on how effective the dust is in obscuring the emerging light (e.g., the Cardiff Meeting, [12]). What makes dust elusive in galaxies is the lack of obvious emission/absorption features, the potential confusion between dust reddening and the aging of the stellar population, and the gray 'net' obscuration often produced by the combination of dust scattering and geometry [43]. Because of these reasons, indirect methods must be usually employed to determine the opacity of a galaxy. Multiwavelength observations to characterize selective obscuration [31, 3, 8, 7], variations of the galaxy opacity with inclination [17, 6], and extinction of background sources by the foreground galaxy [42, 1, 45, 18] are among the most common methods. ⁽¹⁾

The central regions and the arms of spiral galaxies are likely to be opaque, with A_B 1 or larger, while the interarm regions are generally transparent, with A_B 0.3 [42, 17, 1]. Dust may be present in the haloes of galaxies (A_B 0.1, [45]). Active star formation (SF) is usually associated with strong far-infrared emission from the dust heated by the massive stars [35, 37, 21]. The effects of dust in star-forming regions are the result of two competitive processes; on the one side, massive stars are born in the dusty environments of molecular clouds; on the other side, an evolving stellar population tends to blow away or destroy the surrounding dust through supernovae explosions and massive star winds.

Evidence exists for presence of dust at high redshifts. Damped Ly- Systems (DLAs) around z 2-3 have metallicities around 1/10-1/15 solar and dust/gas around 3-20% of the Milky Way value [30, 33]. The DLAs may not be representative of the high redshift galaxy population as a whole, but if so (e.g., [44, 34]), the metal abundances at z=3 are not primordial, and dust, which comes with metals, may be a concern (see, however, Pettini & Bowen 1997). In a Universe which is 1/6-1/4 its present age, only a relatively small fraction of gas has been locked up in stars; gas column densities are larger than in the Local Universe and even a small dust/gas ratio can imply a measurable reddening. In the redshift range z ~ 2.5-3.5, the rest-frame UV is shifted to optical wavelengths; for instance, a spectrum in the wavelength range 4,000-10,000 Å corresponds to the rest-frame range 1,000-2,500 Å of a z = 3 galaxy. Therefore, standard ground-based observational techniques are sensitive to the rest-frame UV emission from distant galaxies, namely to a wavelength region potentially heavily impacted by dust reddening.

¹ Editor's note: See also Trewhella in these Proceedings. Back.