5 Applications of the Synthetic Catalog

An extensive data base of radio observations of HII regions has been distilled into the Synthetic Catalog of 1442 objects at a frequency of 2.7 GHz. We consider some uses of this catalog and the associated Master Catalog for studies of individual HII regions over a range of frequencies and for CMB studies.

5.1 Detectability of HII regions

The sensitivity of a particular instrument in Kelvin to an HII region of a given flux density expressed in Jansky depends upon the observing frequency $\nu$, the beamwidth (FWHM), the sensitivity per second of integration and the observing time t (the rms noise decreases with observing time t as $t^{-\frac{1}{2}}$). The rms noise in Jansky per second of integration, rms_1s,f, is related to the rms antenna temperature in Kelvin per second of integration, rms_1s,a, by:

$$({\rm rms}_{1s,f}/{\rm Jy}) = 2.95 \times 10^{-3} ({\textit FWHM}/{\rm arcmin})^{2} \cdot$$ (1)

$${} (\nu/{\rm GHz})^2 ({\rm rms}_{1s,a}/{\rm K})$$

and the same relation holds between the source signal expressed in terms of flux density or of antenna temperature. Here the source is assumed to be point-like and observed at the centre of a Gaussian, symmetric beam. Eq. (1) can be then used to determine the S/N for a point source of a given flux density. In cases when a source with a diameter comparable or bigger than the FWHM of the observing beam is considered, a correcting factor to Eq. (1) is required to compute the S/N. In particular, the S/N is $\sim$40 to 90$\%$ the value computed with Eq. (1) when a source with a diameter $\sim$2-3 times the beamwidth or $\sim$1/2-1/3 the beamwidth, respectively, is considered. The S/N represents an easy way to compare the signal produced by Galactic HII regions with the sensitivity of a typical CMB anisotropy experiment. Therefore, we apply the above treatment to the high resolution satellite experiment PLANCK by ESA, scheduled for launch in 2007. PLANCK will observe the entire sky with a sensitivity at the end of the mission of about 10 $\mu$K per resolution element. The beamwidths vary between ${\sim} 33.6'$ and 5' from 30 to 857 GHz respectively. For numerical estimates, we consider here the channels at 30 and 100 GHz which have a nominal FWHM of $\sim$33.6' and 10'. For comparison, we make the same calculation for COBE-DMR for which, at 31.5 and 90 GHz the rms temperature was 150 and 100 $\mu$K and the beamwidth was 7$^\circ$ (Boggess et al. 1992).

 

Table 3: Summary of relative (%) errors on quoted fluxes and diameters in the corresponding sub-catalogs. Errors marked by * have been estimated by the authors of the present paper. Errors marked by ** correspond to the indicative diameters estimated by means of the flux-size correlation in Fig. 2. Further details in Sects. 3.1 and 3.2.

Reference	$\%$ flux error	$\%$ size error
Altenhoff et al. (1970)a	10/30	6.5-0.4$\times$$\Theta$ (*)
Altenhoff et al. (1970)b	10/30	33.9+2.5$\times$$\Theta$ (*)
Altenhoff et al. (1979)	5	10
Beard (1966)	33.5-0.1$\times S$ (*)	300 (**)
Beard $\&$ Kerr (1969)	33.5-0.1$\times S$ (*)	37-0.3$\times$$\Theta$ (*)
Beard et al. (1969)	33.5-0.1$\times S$ (*)	37-0.3$\times$$\Theta$ (*)
Berlin et al. (1985)	10	72.5 (*)
Caswell $\&$ Haynes (1987)	10	22.1+2.8$\times$$\Theta$ (*)
Day et al. (1969)	33.5-0.1$\times S$ (*)	300 (**)
Day et al. (1970)	33.5-0.1$\times S$ (*)	300 (**)
Downes et al. (1980)	5	10
Felli $\&$ Churchwell (1972)	15/35	77.4 (*)
Fürst et al. (1987)	10	300 (**)
Goss $\&$ Day (1970)	33.5-0.1$\times S$ (*)	37-0.3$\times$$\Theta$ (*)
Kuchar $\&$ Clark (1997)	10/20	16/25
Mezger $\&$ Henderson (1967)	23.3(*)	20/50
Reich et al. (1986)	10	300 (**)
Reifenstein et al. (1970)	15	3
Thomas $\&$ Day (1969a)	33.5-0.1$\times S$ (*)	300 (**)
Thomas $\&$ Day (1969b)	33.5-0.1$\times S$ (*)	300 (**)
Wendker (1970)	6	-73.6+17.6$\times$$\Theta$-0.5$\times$$\Theta^2$
Wilson et al. (1970)	15	5
Wink et al. (1982)	5/15	73.5-52.1$\times$$\Theta$+11.1$\times$$\Theta^2$ (*)
Wink et al. (1983)	30.3-1.9$\times S$+0.1$\times$ S2 (*)	85.8-38.1$\times$$\Theta$+5.2$\times$$\Theta^2$ (*)

The S/N for all the HII regions in the Synthetic Catalog have been calculated as discussed above; the flux densities were estimated at each frequency from 2.7 GHz values assuming a flux density proportional to $\nu^{-0.1}$. Figure 7 shows the number of sources per bin of S/N (the distribution function) and number of sources greater than a certain S/N (cumulative distribution function) for both PLANCK and COBE-DMR at $\sim$30 and $\sim$100 GHz.

Figure 7: Distribution functions (bottom curves) and cumulative distribution functions (upper curves) at 30 and 100 GHz for PLANCK (dotted lines) and COBE-DMR (dash-dot lines). The COBE-DMR angular resolution is 7$^\circ$ at both frequencies.

From Fig. 7 it is clear that the vast majority of sources in the Synthetic Catalog produces a signal which highly exceeds the detection threshold of the instrument. Therefore, not only all the Synthetic Catalog HII regions will be detected but also $80\%$ of these sources will have a S/N > 100 (while the weakest sources, with $S \sim10$ mJy, will be seen with a $S/N \sim 3$). As a consequence, PLANCK high sensitivity and high resolution should allow to significantly entend the existing HII regions data base. It is important to emphasize that none of the individual HII regions would have been seen by COBE-DMR whose flux sensitivity according to Eq. (1) is $6 \times 10^{3}$ less than PLANCK at 30 GHz and $4 \times 10^{4}$ less at 100 GHz.

Figure 8: Simulated full sky map of the antenna temperature signal at 30 GHz produced by the Synthetic Catalog HII regions for an instrument with a FWHM of 33.6'. The map is displayed in all-sky mollweide projection. The color bar shows the minimun and maximum of the pixels in the map with a positive signal. The signal looks particularly bright with peaks of tens of mK. See also the text.

Figure 9: Simulated maps of the Cygnus X region (left panel) and of the Galactic Center region (right panel) at 30 GHz with a FWHM resolution of 33.6'. The units of the color bar are mK. Minimum and maximum refer to positive pixel signal values.

5.2 Radio/millimeter studies of HII regions

The ionizing radiation from the central O/B stars produces the HII region which emits through the free-free mechanism from radio to submillimetric wavelenghts; the surrounding dust is also heated by radiation originally derived from the central stars and as a consequence radiates at submillimeter and IR wavelenghts. The flux density of the radiation from these two components was found to be equal in the wavelenght range 1-3 mm (100 to 300 GHz) by Kurtz et al. (1994) for a selection of compact HII regions. However, many interesting questions remain to be resolved in the physical relationship between the HII region and the radiation-heated dust cloud in which it is embedded. The Synthetic Catalog provides a rich source of HII regions for further study. Figure 7 shows that many hundreds of HII regions will be detectable at high sensitivity and with adequate resolution over the frequency range 30 to 857 GHz by PLANCK. Thus, this may represent a good chance for a comparative study of the far-IR and radio continuum morphology of Galactic HII regions.

Moreover, source identification in IR experiments such as DENIS (Epchetein et al. 1994) and 2MASS (Kleinmann 1992) can benefit from the crosscheck with the Synthetic Catalog. In fact, for these kind of experiments a major problem is the association of an observed source with a bright IR Galactic object like an HII region or an ultracompact HII region rather than with an external Galaxy.

5.3 Use of the Synthetic Catalog in CMB studies

We will consider in this subsection several applications of the Synthetic Catalog in CMB studies. The first use we discuss is the production of maps of the integrated free-free emission as seen with the angular resolution of an instrument such as PLANCK at each observing frequency. Free-free emission dominates the Galactic plane signal at least over the range 30 to 100 GHz. Figure 8 shows the Galactic plane emission resulting from HII regions as seen by PLANCK at 30 GHz where the beamwidth is 33.6'.

To make the map in Fig. 8, we have implemented a code which simulates the contribution of each source in the Synthetic Catalog at a given angular resolution by assuming a symmetric Gaussian profile for the source brightness distribution and numerically convolves, in real space, the relevant part of the map - obtained after considering the contribution from all the sources in the catalog - with another symmetric Gaussian profile having the instrument FWHM. The final map is generated in HEALPix (Hierarchical Equal Area and IsoLatitude Pixelization of the Sphere, by Górski et al. 1998). Figure 9 shows the strong Galactic centre and Cygnus X regions. Antenna temperature as high as $\sim$50 mK are seen.

Moreover, HII regions, as delineated in Fig. 8, can in principle represent a significant contribution - comparable to that produced by synchrotron and dust emission - to one of the most critical spurious effects in CMB surveys, the so-called straylight contamination through their integrated signal in the sidelobes (Burigana et al. 2001) which needs to be properly evaluated. The optical design of PLANCK and similar mapping instruments must be optimized to minimize this effect. In any case, the residual straylight should be taken into account in the data analysis.

HII regions also have a part to play in CMB imaging experiments as suitable calibrators and as probes of the pointing and beamshape. HII regions are non-variable and have quite a well-known spectrum which makes them valuable calibrators along with planets and the CMB dipole (Bersanelli et al. 1997). In addition, for space missions, they provide auxiliary data for inflight beam reconstruction and pointing by complementing the information from planet transits (Burigana et al. 2002) and from the interplay between amplitudes and phases of CMB signal with the instrumental noise (Chiang et al. 2002). In particular, the Chiang et al. method, although not requiring the use of bright sources, allows the reconstruction of the beam ellipticity only in the beam central regions while the Burigana et al. technique allows the complete reconstruction of the beam shape down to a level of -25 dB but makes use of non-variable, bright, compact sources. However, despite being originally conceived for planets, the Burigana et al. method can be easily extended to other classes of sources which also enable to increase the number of transits over the space mission lifetime. The Synthetic Catalog contains 36 HII regions which have a flux density $\ge$30 Jy at 30 GHz and a diameter $\le$ than 5'. Accurate flux densities ad positions can be determined from ground-based aperture synthetis observations.

Finally, we point out that the angular extension of a typical Galactic HII region represents an intermediate case between point sources and diffuse foregrounds for which component separation tools have been specifically designed (see, e.g., Maino et al. 2001, and references therein). The fluctuations produced by such extended sources and the capability of existing component separation methods to handle with them will have to be furtherly investigated in the next years.