CFHT Capabilities for Mid-Infrared Observations:
A Bonus for ISO

Isabelle Vauglin

Observatoire de Lyon
Electronic-mail: vauglin@obs.univ-lyon1.fr

Christine Joblin

CESR/CNRS Toulouse
Electronic-mail: joblin@cesr.fr

and

Philippe Merlin, Jean-Pierre Dubois

Observatoire de Lyon

Abstract:

The contribution of the ISO satellite to the advances of infrared astronomy will certainly be tremendous. However, ground-based mid-infrared astronomy can still bring a lot, especially if one disposes of a high-quality site. Thanks to the excellence of the CFHT site on Mauna Kea, the 10 micron camera C10$\mu$ proved to be a powerful instrument to complete ISO observations by combining high spatial resolution and very good sensitivity.

Introduction

  CFHT is known to be a very good site for infrared observations, and perhaps the best ground-based site for observations in the thermal infrared domain (Vauglin and Merlin, 1995). The infrared camera C10$\mu$ has been developed by the Observatoire de Lyon to work in the 5 - 15 $\mu$m range. The camera was upgraded in 1996 with an Si:Ga array of 128x192 pixels (from LETI-LIR, Centre d'Etudes Nucléaires de Grenoble, France). The instrument is a common-user instrument on CFHT (Merlin, Sibille and Vauglin, 1996).

The results which are presented in this paper are based on data obtained at CFHT with C10$\mu$ in July 1997 by C. Joblin and the Lyon team. The aim of the programme was to complement in the 8-13 $\mu$m window the observations of extended objects (photodissociation regions) obtained with ISOCAM. The ground-based observations were further motivated by the focal scale of 0.3 arcsec/pixel offered by C10$\mu$.With this high spatial resolution, structures of the diffuse medium matter at scales much smaller than those observed by ISOCAM (3 arcsec resolution at best) are expected to be revealed. Results and new sensitivity limits for C10$\mu$ are presented.

Ground-Based Observations in Thermal Infrared

 

Difficulties

  Observations from the ground in the 5 to 15 $\mu$m range are difficult, for essentially two reasons:

• The atmospheric window: The atmosphere is only partially clear is this domain, with an almost complete opacity from 5.5 to 7.7 $\mu$m and from 13.9 to 16 $\mu$m due to the absorption by H₂O and CO₂ respectively. The quality of the atmospheric transmission greatly improves with the altitude and the dryness of the site.
• The background emission: The peak emission of a 300 K black body occurs precisely at 10 $\mu$m. The huge emission which arises from the immediate environment and the atmosphere (background emission) makes the observation of an astronomical source comparable to "the search of a needle in a haystack". The signal from the source is lost in the background emission and the observations are quite always photon-noise limited. In addition, the background emission is variable on a short time scale (< 100 msec). This imposes a continuous measurement and subtraction of the sky. The integration times are thus very short (16 to 60 msec) and both nodding and chopping are essential.

The sensitivity limits which can be reached in a given site are therefore directly linked to the intrinsic quality of this site: transparency, level and stability of the background. This requires a high altitude site and a dry atmosphere. Because of these exacting conditions, there is a crucial need for a tailored site. In that sense, CFHT is outstanding for thermal infrared observations. It must be noted that a site cannot be built or even improved. We can only match instruments to its qualities. Mid-infrared observations are particularly constraining, but fortunately CFHT is exceptionally good for the 10 $\mu$m window. We have to take advantage of the fact!

Advantages

Despite the encountered difficulties, observing from the ground has several advantages:

• The spatial resolution: At 10 $\mu$m and on a 4 meter telescope, the size of the diffraction disc is 0.6 arcsec. With a pixel scale of 0.3 arcsec, the observations are diffraction limited and the diffraction disc is correctly sampled. With its 0.6 meter telescope, ISO can only provide a spatial resolution of 3 arcsec. This is 5 times less than the expected resolution for the C10$\mu$ instrument at CFHT.
  
• The time limitation: Observations from the ground are more flexible than space observations. In particular, observations can be done several times, and longer integration times are in general allocated. Furthermore, larger collecting areas are available and the instruments can benefit from technical improvements. For example, the first detector of C10$\mu$ was a 32x32 array identical to the long-wavelength channel array of ISOCAM. It has been changed twice since that time and disposes now of a larger array (128x192 pixels) of better quality.
  
• The availability of CFHT for IR astronomy: The quality of the CFHT site makes it ideally suited for mid-IR observations from the ground. The sensitivity limits which have been obtained (cf. section 3) definitely prove that this site is ideal for many ISO follow-ups.
  

C10$\mu$: Description and Detection Limits

  C10$\mu$ is optimised for ground-based observations in the thermal infrared. Its array has benefited from the progresses made by LETI-LIR (CENG, France) in the technology of these detectors. With a sampling of 0.3 arcsec/pixel, the 128x192 pixel array has a total field of view of 40x60 arcsec. A broad-band N filter and nine different narrow-band filters are available, as well as two CVFs, one for the 4.4 $\mu$m-7.85 $\mu$m domain, the other for the 7.72 $\mu$m-13.64 $\mu$m domain. They have a spectral resolution of about 50. The shortest integration time is 16.6 msec; it can increase to more than 100 msec, if the conditions allow so. Generally, the observations are limited by the photon noise.

During the C10$\mu$ run of July 97 at the CFHT, the weather conditions were excellent with 0% humidity. The detection limits reached during this run are remarkable for a ground-based instrument. A signal to noise ratio of 10 was obtained in 1 hour of integration time on extended sources for the following flux:

- 39 mJy in the PAH filter (11.23 - 11.48 $\mu$m)
- 26 mJy in the 8.37 - 8.88 $\mu$m filter (comparable to the filter LW2 of ISO)
- 21 mJy in the silicate filter (9.17 - 10.09 $\mu$m)
- 28 mJy in the 11.85 - 13.1 $\mu$m filter.

Observations

 

Scientific justification

One of the important question of the mid-IR astronomy is the nature of the carriers of the IR emission bands at 3.3, 6.2, 7.7, 8.6 and 11.3 $\mu$m. These bands, usually called the unidentified IR (UIR) bands, are the signature of aromatic C-C and C-H bonds. The main candidates are polycyclic aromatic hydrocarbons (PAHs; Léger and Puget 1984) and various forms of hydrogenated carbon grains (Borghesi et al. 1987, Papoular et al. 1989). The spectra of PAHs, measured in the laboratory, have not provided a fully convincing match of the observed spectrum yet. On the other hand, no emission mechanism has been found to account for the emission of larger grains. The emitting species are, therefore, probably intermediate compounds containing hundreds of carbon atoms (Joblin et al. 1997, Joblin 1998a).

The ISO satellite has provided a lot of spectral information on the UIR bands. This data is expected to further constrain the chemical identity of the carriers of these bands. For example, ISO showed the similarity of the observed spectrum in regions of the interstellar medium where the UV flux differs by orders of magnitude: for example in the diffuse ISM (Mattila et al. 1996) and in the photodissociation region (PDR) associated with the HII region in M17 SW where the UV flux is 10⁵ times higher (Cesarsky et al. 1996, Verstraete et al. 1996). Despite the global similarity of the spectrum, variations of the relative band intensities within extended objects have been reported by several authors (for instance on M17 by Cesarsky et al. 1996, and Verstraete et al. 1996). The observed variations have been attributed, in the PAH model, to changes in the size distribution or/and in the ionisation degree and hydrogenation coverage of PAHs. However, the studied regions have generally complex geometries which induce local variations of the density and of the excitation conditions. A spectrum measured in a large beam may just represent the average of extreme values (beam of 14"x10" for the SWS measurements of Verstraete et al., pixel size of 6"x6" for the ISOCAM maps of Cesarsky et al.). Detailed studies at higher spatial resolution are clearly required to study the spatial distribution of the UIR bands in relation to the physico-chemical conditions and to the nature of the band carriers.

The observations presented here are part of a programme which combines ground-based and ISO observations. The observed regions are photodissociation regions (PDRs) associated to HII regions corresponding to the following objects: S106, IRAS 19442+2427, and M17 SW. In this paper, we present some of the obtained images which illustrate the capabilities of the C10$\mu$ instrument. The scientific interpretation of this data will be reported elsewhere.

Sharpless 106

The massive Young Stellar Object S106 was chosen because of its extension and brightness. Strong variations of the local physical and chemical conditions are also present within this object (Gehrz et al. 1982). A first interesting result which was obtained is the comparison of the flux measured by C10$\mu$ in the 8.6 $\mu$m UIR band with that measured by ISOCAM. The CVF observations of ISOCAM were integrated in the 8.6 $\mu$m filter of C10$\mu$ and C10$\mu$ images were degraded to the spatial resolution of 3 arcsec/pixel. The comparison of the composite ISO image and the degraded C10$\mu$ image clearly underlines the still-poorly characterised straylight problem in ISOCAM images which leads to an increase of the flux. The detailed comparison can be foundin Joblin et al. (1998b). Figure 1 illustrates the complicated structure of the matter in the S106 object as revealed by the C10$\mu$ image at 8.6 $\mu$m. For comparaison, Figure 1 also shows the ISO map at the same wavelength.

 

Figure 1:   Comparison of the C10$\mu$ and the ISOCAM images of S106 measured in the 8.6 $\mu$m PAH band with pixels of 0.3 and 3" respectively.

The spatial distribution of the various UIR bands within the S106 object will be discussed in a forthcoming paper (Joblin et al. 1998c). As an example, Figure 2 shows the emission at 11.3 $\mu$m and in the adjacent continuum.

 

Figure 2:   C10$\mu$ images of S106 in the 11.3 $\mu$m emission band (top) and the adjacent continuum (bottom). Contours are at the same levels for both maps. The emission in the band is obviously more extended than in the continuum.

Other sources

Other sources observed with C10$\mu$ in July 97 are M17 Southwest and IRAS 19442+2427 in narrow-band filters, and NGC 7027 in CVF mode. The object IRAS 19442+2427 is not well-known in the literature. Its mid-IR spectrum was first reported by Jourdain de Muizon et al. (1990). The central source happens to be double, as revealed by the C10$\mu$image (Figure 3). The angular separation is measured to be 1.6 arcsec. Further studies on this object combining ISO measurements are now in progress.

 

Figure 3:   C10$\mu$ observations of IRAS 19442+2427 in PAH filter (11.3 $\mu$m). The central source is clearly double.

Conclusion

In the mid-infrared, the background is so high and so variable that one must dispose of a particularly well-adapted site and have a very clean and strict observing procedure. Only the optimal combination of these two points can lead to the level of detection reached by C10$\mu$on the CFHT. The CFHT community disposes then of an instrument which is particularly well suited to undertake ISO follow-ups. Further than providing complementary data (maps at different wavelengths, cross-check calibrations...), C10$\mu$can bring new insights into the ISO science by providing data at much higher spatial resolution (0.6 arcsec).