5 The Galactic foreground extinction

A detailed analysis of the Galactic foreground extinction is presented by Woudt (1998). Here, the two most common extinction indicators are briefly discussed.

5.1 The neutral hydrogen in the Milky Way

Assuming no variation in the gas-to-dust ratio, the neutral hydrogen (HI) content in the Milky Way can be used as an indicator of the foreground extinction. Close to the plane of the Milky Way (within $1{{^\circ}}{-}2{^\circ}$ of the Galactic Equator), the Galactic HI line might be saturated, thereby underestimating the true extinction. At these latitudes the Galactic CO (Dame et al. 1987) might be a better tracer. However, as we have found no galaxy candidates that close to the Galactic Plane, the HI column densities would in principle be adequate.

Following the precepts of Burstein & Heiles (1978, 1982), the Galactic foreground extinction A_B = $1.337 \times A_V = 4.14 \times E_{(B-V)}$ (Cardelli et al. 1989) can be determined from the Galactic HI column density alone by

$$E_{(B-V)} = \left({{N({\rm HI})}\over{2.23 \times 10^{18}}}\right) \times 4.43 \times 10^{-4} - 0.055 \\$$ (1)

where $N({\rm HI})$ is in units of 10²¹ atoms cm^-2. However, the gas-to-dust ratio does vary locally. As discussed in Woudt (1998), we support the earlier finding by Burstein et al. (1987), that the HI gas-to-dust in the Crux and Great Attractor region is twice the nominal value. A higher gas-to-dust ratio locally implies an overestimation of the Galactic extinction.

5.2 The 100 micron DIRBE/IRAS extinction maps

Schlegel et al. (1998) have presented the 100 micron extinction maps from the DIRBE experiment. These maps have a much better angular resolution (6 $\hbox{$.\mkern-4mu^\prime$ }$1) compared to the HI maps ($\sim$20-30 arcmin) and variations in the gas-to-dust ratio are not important. These maps provide a direct measure of the dust column density. The reddening maps, however, have only been calibrated by the colours of elliptical galaxies ((B-R) and (B-V), see Schlegel et al. 1998 for details) at intermediate and high Galactic latitude.

Woudt (1998) has tested the calibration of the reddening maps at low latitude using the (B-R) colour and Mg₂ index of 18 elliptical galaxies in the Crux and Great Attractor region. The agreement between the reddening values of the DIRBE/IRAS extinction maps and the E_(B-V)values derived from the colour-Mg₂ relation (Bender et al. 1993) is generally good (Woudt 1998). Note, however, that because of the choice of B and R filters the latter values are very sensitive to the assumed Galactic extinction law (R_V = 3.1). Large systematic uncertainties could therefore arise when deviations from the mean Galactic extinction law occur. A different choice of filters, e.g. R and K', eliminates this uncertainty because the Galactic extinction law is insensitive to the assumed value of R_V in the near infrared (see Cardelli et al. 1989). A calibration of the reddening maps at low Galactic latitude using the (R-K') colour of low-latitude elliptical galaxies is highly desirable. For more details on the problem of selecting the "right'' R_V value, see McCall & Armour (2000).

In this paper, we exclusively use the DIRBE/IRAS reddening maps for extinction corrections. Furthermore, we assume R_V = 3.1 for the conversion of selective extinction to total extinction, and A_B/A_V = 1.337 (Cardelli et al. 1989).