Astron. Astrophys. 318, L35-L38 (1997)

3. Discussion

The galactic anticenter region, near , is particularly well-suited to a search for the DI hyperfine emission feature at 92-cm. The HI emission brightness in this region peaks smoothly into a plateau extending spatially over several degrees with a brightness temperature 135 3 K, as can be seen in Burton (1985). Conditions should therefore be reasonably constant over the beam of the WSRT dishes. The extended region of line brightening is paired with a line narrowing due to velocity crowding. The HI emission-line profile from this region is relatively narrow, with a FWHM of 18.7 km s^-1. The line core is rather flat-topped, strongly suggesting a high line opacity. It also displays shallow self-absorption, indicating a modest amount of temperature substructure along the line-of-sight. Based on the depth of the self-absorption features and assuming a substantial line opacity, the spin temperature of the atomic gas appears to be in the range 125-135 K.

Direct measurements of the HI opacity along many lines of sight in this region have not yet been carried out. The typical sky density of moderately bright extragalactic background sources suitable for such a measurement suggests that a few tens of lines of sight could, in principle, be observed. The closest line of sight that has been observed in HI absorption (Dickey et al. 1983) is at . This is near the edge of the plateau in emission brightness, where the line profile is already somewhat broader than in the direction of maximum velocity crowding, with a FWHM of 23.4 km s^-1. The absorption profile extends smoothly over the entire velocity extent in which the emission profile exceeds a brightness of about 5-10 K, with a uniform high opacity of about 2 across the line core. This is consistent with the expectation for a gas distribution in which the column density is dominated by an approximately isothermal gas with kinetic temperature in the range 125-135 K. An absorption equivalent width,  = 33.68 km s^-1 is found. The very smooth nature of the HI emission suggests that it is plausible to expect a comparable equivalent width in the direction .

Assuming that the DI emission is either mixed with, or lies behind, most of the galactic continuum emission, (which has the substantial brightness of about 70 K at 92 cm wavelength) the equation of radiative transfer yields simply:

where is the differential brightness temperature on and off the line, is the spin temperature of the DI and is the optical depth of the DI. For this yields , or for an approximately isothermal gas. The line integral of the Gaussian overlaid on the possible feature in Figure 1 is  K km s^-1. Assuming that the spin temperature of the DI is the same as that of HI (i.e. 130 K), this corresponds to  = 3.7 10^-4  km s^-1. Hence the estimate of the ratio of the optical depths is . The relationship between the ratio of the optical depths and the ratio of the column densities of DI and HI is (Anantharamiah & Radhakrishnan 1979) , hence the estimated abundance of DI is . This is comparable to the inferred pre-solar abundance (Kunde et al. 1982, Courtin et al. 1984), and about a factor of 2 above the current DI abundance in the local solar neighborhood (Linsky et al. 1993).

As discussed in the introduction, the lack of detection of DI towards the inner Galaxy suggests that most of the DI has been converted into HD within molecular clouds (Heiles et al. 1993). Towards the outer Galaxy, however, the molecular column density is low along many sight lines. In fact, the survey by Dame et al. (1987) detected very little CO emission in the galactic plane between 180 and 185 degrees longitude except for a small concentration at -10 km s^-1. The high optical depth in HI seems to be due to velocity crowding in diffuse gas along a pathlength of several kpc, and does not appear to arise in a single physical entity like a giant molecular cloud. It seems plausible, therefore, that a large fraction of the deuterium in this direction is still in atomic form. Further, the metallicity gradient in the Galaxy suggests that the DI abundance in the outer Galaxy would be higher than the inner. Calculations by Prantzos (1996) show that the deuterium abundance gradient is in general steeper than (and of course opposite in sign to) the oxygen gradient (because of the late ejection of metal-poor but deuterium-free gas from low mass stars formed at early times). While the exact gradient is sensitive to the assumed infall model, the current DI abundance within diffuse atomic gas in the inner and outer Galaxy could well differ by a factor of two.

Another useful comparison to make is with the range of measured DI abundances, , in high redshift Lyman limit systems (Carswell et al. 1994, Songalia et al. 1994, Tytler et al. 1996). If a high primordial abundance turns out to be correct, our results indicate that even in regions with reduced star formation, the current deuterium abundance is about a factor of 6 lower than primordial. In contrast, recent model calculations of astration of deuterium suggest that the abundance evolution is modest, with current abundance only a factor less than primordial (Galli et al. 1995). On the other hand, the low value of D/H measured by Tytler et al. (1996), , is consistent with our own possible detection and the inferred pre-solar value. Direct imaging of DI emission in the outer regions of nearby galaxies may provide a very effective means of addressing this issue comprehensively, once the capability for achieving sub-mK sensitivity on arcmin scales becomes available with the next generation of radio telescopes. The Giant Meterwave Radio Telescope (GMRT) should already allow a robust detection of DI emission from the Galaxy within the next few years, while DI imaging of nearby external galaxies should become possible with construction of the Square Kilometer Array Interferometer (SKAI) early in the mext millenium.

© European Southern Observatory (ESO) 1997

Online publication: July 8, 1998
helpdesk.link@springer.de