6 Correlation with dust emission

Both FIR continuum emission from dust and CO spectral line emission can be used to estimate the masses of interstellar clouds, and the FIR data can be compared for consistency and/or calibration with the CO data. For isolated clouds, spatial coincidence and close morphological similarity is found between the CO emission and dust emission, especially after subtraction of the extended background Galactic FIR emission (Heyer et al. 1987; Langer et al. 1989; and Mooney 1992). In general, an accurate separation of the background Galactic FIR emission from the cloud emission is necessary before using the IRAS data to estimate the FIR luminosity of a cloud. However, for the region in this study the identification of the FIR emission is fairly straightforward as the cloud is well isolated. The only other contribution to the FIR emission would be atomic hydrogen. An HI map was available from the Weaver & Williams (1973) data base, although the resolution of the original data is very large, 36'. We have resampled it onto a 3' grid, assuming there is no specific structure. In Fig. 9 we present the HI integrated intensity map fitted to the CO mapped region. The distribution of HI emission is found to have a very small dynamic range of a factor of 1.3, whereas the CO emission without internal heating sources has a much larger dynamic range. Thus the atomic hydrogen does not affect significantly the slope of the relationship between the FIR emission and the CO emission; we have not included the HI contribution in this work.

  

Figure 10: The 100 $\mu$m intensity as a function of CO integrated intensity. Points with I100 less than 35 MJy sr-1 were fitted using the least-squares technique and the fitted line is shown. The points that lie substantially above the fitted line larger than 35 MJy sr-1 are due to IRAS point sources and Sh 241

Figure 10 represents the pixel to pixel correlation between the CO integrated intensity and dust FIR intensity at 100 $\mu$m. It shows two distinctly different features; a very scattered distribution of FIR intensity above 35 MJy sr^-1, and the bulk of emission confined to between 15 and 30 MJy sr^-1. However, we note that the pixels with FIR intensity greater than 35 MJy sr^-1 are all associated with the HII region Sh 241 and IRAS point sources. Except for those points, a good correlation is found between the FIR emission and the ¹²CO integrated intensity. To determine the relation between ¹²CO integrated intensity and FIR intensity, pixels with the CO integrated intensity > 5 K km s^-1 ($3\sigma$) have been fit by using the least-squares method. The slope of the correlation is 0.38 (MJy sr^-1) (K km s^-1)^-1 and the intercept is 21.0 MJy sr^-1, which is represented as a thick solid line in Fig. 10. The non-zero intercept most likely represents emission from dust in a region which lacks ¹²CO. The amount of dust that lies in front of or behind the target region may not be small, as the intercept of above equation is nonnegligible. HI contribution to the FIR emission is also involved as mentioned previously. However, considering that it is the general dust emission from the Galaxy, thus not associated directly with the CO integrated intensity, only the slope represents the true relation between the ¹²CO integrated intensity and the FIR intensity. The good correlation implies that there is little stray dust emission in the line of sight.

Numerous studies have found a linear relationship between 100 $\mu$m intensity and either the HI column density in regions of atomic gas or the CO column density in molecular clouds (Boulanger & Pérault 1988; Snell et al. 1989; Laureijs et al. 1995; Boulanger et al. 1998). Boulanger & Pérault (1988) found ratios between the 100 $\mu$m intensity and CO integrated intensity in the range 0.6 to 2.5 (MJy sr^-1) (K km s^-1)^-1 with an average value of 1.4 (MJy sr^-1) (K km s^-1)^-1 for regions outside of star forming sites. This value can be compared with the average slope found for the target region of 0.38 (MJy sr^-1) (K km s^-1)^-1, substantially less than this value. Based on the slope of the least-squares fit to the data, we have computed the ratio of 100 $\mu$m intensity to total hydrogen column density. We can use the CO conversion factor computed from the $\gamma$-ray analysis (Bloemen 1989), of $2.3 \ 10^{20}$ H₂ ${\rm cm^{-2}\ (K\ km\ s^{-1})^{-1}}$ to convert the CO intensities to hydrogen column densities. Expressing this ratio in terms of the hydrogen, the values found for the target region are $I_{\rm 100\ \mu m}/N$(H) = 0.07 (MJy sr^-1) (10²⁰ H cm^-2)^-1. These values can be compared to those for dark clouds; $I_{\rm 100\ \mu m}/N$(H) is 0.07 (MJy sr^-1)(10²⁰ H cm^-2)^-1) for B18, and 0.10 (MJy sr^-1)(10²⁰ H cm^-2)^-1 for HCL2 (Snell et al. 1989). The values for warmer GMCs are substantially higher; for example, in Orion Molecular Cloud (OMC): $I_{\rm 100\ \mu m}/N$(H) = 1.3 (MJy sr^-1)/(10²⁰ H cm^-2)^-1 (Boulanger & Pérault 1988). Thus, the ratios found for the target region are comparable to the dark cloud ratios, and a factor of $\sim$ 20 times smaller than the ratios in OMC. Snell et al. (1989) attributed the low ratios in the dark clouds to dust heated exclusively by the solar neighborhood interstellar radiation field. A similar conclusion was reached by Mooney (1992) for the clouds classified as IR-quiet. Thus, the low ratios found for the target region probably results from the absence of internal or significant nearby external heating sources. If this is the case, the dust temperature should also be relatively low as is seen in the dark clouds.