Astron. Astrophys. 318, 608-620 (1997)

2. Observations

2.1. ESO

We obtained a series of infrared images of OMC-1 during two observing runs from Dec 29, 1993 to Jan 1, 1994 and Dec 24 to 27, 1994 with the ESO/MPI 2.2m Telescope and the infrared camera IRAC2.

A Fabry-Pérot filter was centered at various K band transitions of the H₂ molecule and the H molecular ion. The bandwidth of the FP was 1.9 nm and the accuracy of the wavelength setting was 0.5 nm. The detector was a 256 256 NICMOS array with a field of view of about . The detector pixel projected to . Integration times were 5 min on source and 5 min on two separate background positions. For weaker lines, several cycles were averaged. The seeing conditions during the first observing run were excellent and the measured FWHM of the stellar images was . The second run had slightly worse seeing conditions and in addition suffered from an RA tracking problem. Some of the images are thus slightly elongated in the east-west direction.

The H₂ transitions observed were S(0) 1-0, S(1) 1-0, S(1) 2-1, S(2) 2-1, S(3) 2-1, S(3) 3-2, S(3) 4-3, Q(1) 1-0 and O(5) 8-6. The S(1) 2-1 at 2.2477 µm image is contaminated by emission from M42. Possibly this leakage is due to [Fe III ] radiation. An image of OMC-1 taken at 2.122 µm is shown in Fig. 1 and a continuum subtracted and deconvolved image in Fig. 2. Continuum images were obtained at 2.1319 µm, 2.1577 µm, 2.2098 µm, 2.2334 µm and 2.3652 µm. The continuum was found to be constant within the observational errors and the frames with the best seeing images were averaged. Image deconvolution (Fig. 2) was achieved with the Lucy algorithm (Lucy 1974) after 20 iterations and a bright star in the field as PSF.

Fig. 1. Inner region of OMC-1 at a wavelength of 2.121 µm. The Becklin-Neugebauer object is marked with a black cross. The white circle shows the position of the compact radio source I which lies close to IRc2 (Menten & Reid 1995). The Trapezium stars are visible at the bottom. Coordinates are for J2000.0.

Fig. 2. Continuum subtracted and deconvolved version of the above image. The horizontal bar represents .

The transitions used for the search for the molecular ion H were those in the overtone band used by Drossart et al. (1989) in their original detection of H on Jupiter. Specifically we used R(6) at 2.0933 µm, Q(4) at 2.1944 µm, R(8) at 2.1342 µm, P(5) at 2.2028 µm and P(3) at 2.2039 µm. Continuum images were taken at 2.0948 µm, 2.1330 µm, 2.1370 µm, 2.1924 µm, 2.1934 µm, 2.1954 µm, 2.1964 µm, 2.2052 µm.

Since that Fabry-Pérot filter is uncooled, there are strong thermal emission rings. Although these can be subtracted with proper sky frames, detection of weak emission towards the detector edge is hampered. Pairs of flatfield frames were taken from a dome screen with the lamp on and off such that the thermal FP emission could also be removed in the flatfields. The Fabry-Pérot images were flux calibrated with stars taken from McCaughrean and Stauffer (1994). Typically we have about a dozen stars in the southern third of our frames in common. The RMS scattering of the local sensitivities thus obtained at the position of each calibration star was less than 10 %. We estimate that the accuracy of the absolute intensity calibration across the area of OMC-1 is better than 50 % and that on small scales ( ) it is better than 20 %.

2.2. UKIRT

We observed the Peak 2 region in OMC-1 spectroscopically during the nights from Sep 29 to Oct,1 1993 with the UK infrared Telescope and the facility spectrometer CGS4. The detector was an InSb array with 58   62 pixels. A K band spectrum was obtained with the 75 lines/mm grating and the 300 mm camera. It covered the wavelength range from 1.99 to 2.20 µm with a resolution of R 1400. The slit length was and a pixel projected to on the sky. Exposure times were 150 sec, both on sky and target. A relative flux calibration was obtained with observations of the B8 star BS 1713.

With the Echelle grating and the same camera a spectrum centered at 4.0 µm was obtained. This configuration provides a nominal dispersion of 19 km/sec per pixel. The resolution measured from the argon arc is R 9000. Because of anamorphic magnification in this configuration the pixel projects to , in the spatial and spectral directions, respectively. Sky background exposures were taken at offsets of . Total integration times on the sky and target were 300 sec each. For flux calibration the standard star BS 788 was observed. The slit was orientated east-west for both observations and its position on OMC-1 is shown in Fig. 6.

A further spectrum at the same sky position was obtained during the night of Jan 29, 1994 as a service observation by Tom Geballe and John Davies. The 150 l/mm grating together with the short (150 mm) camera provided a wavelength coverage from 3.9 to 4.1 µm with a resolution of R 1200. The same detector was used but with this camera the pixel size corresponds to on the sky. The total exposure time was 84 sec on target and sky each.

As an illustration we show the observed spectra at the strongest emission peak in Figs. 3 & 4. In the K window (Fig. 3) lines of molecular hydrogen dominate the spectrum. There are also recombination lines of hydrogen and helium from the Orion Nebula M 42. These lines have a completely different spatial profile such that in the case of both, strong and weak lines, a clear separation between molecular (OMC-1) and atomic (M 42) origin is possible.

Fig. 3. K band spectrum taken at the brightest knot of Peak 2 (see Fig. 6). Transitions of molecular hydrogen are marked with their spectroscopic designation. The lines of hydrogen and helium are dominated by emission from M42. The observational errors are comparable to the line thickness.

Fig. 4. High resolution spectrum around 4 µm of the brightest knot in the Peak 2 area (see Fig. 6). The lines of molecular hydrogen (S) and atomic species are indicated.

In the spectrum around 4 µm (Fig. 4) the S(12) 0-0 and S(13) 1-1 transitions of molecular hydrogen are clearly detected. The strongest emission line is Br . We note a weak unidentified emission feature at 3.942 µm.

Our Echelle spectrum also covers the S(12) 0-0 line which has its upper level at 15 549 K. Over most of the length of the slit the line is unresolved, i.e. the intrinsic line width is less than 33 km/sec. At the western edge of the bright spot (cf. Fig. 6), the intrinsic line width suddenly increases to about 60 km/sec. Within a separation of only , the turbulent velocity of the molecular hydrogen can therefore increase by a factor 2.

Fig. 5. Wavelet analysis of the H2 S(1) 1-0 emission in OMC-1. Top left: Original continuum subtracted image. The northern jet and one of the south-eastern jets are marked by the arrows (see text). Bottom left: Wavelet transform with the different scale images superposed. Four panels at right: Wavelet transforms at four different scales to enhance structure at small-scales (top center), intermediate-scales (top right and bottom center) and large-scales (bottom right).

Fig. 6. OMC-1 in the light of the H2 transitions S(3) 2-1, S(3) 3-2, O(5) 8-6 and the continuum (taken at 2.098 µm and 2.334 µm). The white line in the top left panel shows the position of the slit of our spectroscopic observations. The H2 images are continuum subtracted.

Our spectral observations cover the wavelengths of several transitions of the molecular ion H . In the K band we cover the transitions R(6) at 2.0933 µm and R(8) at 2.1342 µm. In the range 3.9 - 4.1 µm there is a blend of two Q(3) transitions at 3.987 µm and the Q(1) transition at 3.953 µm (Kao et al. 1991). No lines were detected, however. Our spectra may be used to provide upper limits to the H concentration, which will be derived and discussed below.

© European Southern Observatory (ESO) 1997

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