3. H2 morphology of OMC-1
Fig. 1 shows an observed image of the OMC-1 complex at 2.121 µm. In the north are Beckwith's Peak 5 (which nearly coincides with IRc9) and Peak 1. Near the centre BN shows up brightly. Peak 3 is resolved into two bright sources, south-east of BN. IRc2 lies a few arcseconds north-west of a bright triangular shaped emission area. East-south-east of Peak 3, Peak 2 is a bright region which is resolved into a network of crossing linear structures. A strong emission knot can be seen south-east of BN. The central object in Peak 5 appears to be surrounded by a partial ring in the north-east and south-west. In the neighbourhood of Peak 1, linear strings stretch from south-east to north-west.
3.1. Structural analysis
The continuum subtracted image at 2.121 µm (Fig. 5, top left) shows that the S(1) 1-0 emission of molecular hydrogen is mainly due to a superposition of near-linear features. Without wishing to prejudice the discussion below on possible mechanisms for the formation of the structure observed, we will follow Allen & Burton (1993) in calling these "jets". They seem to expand in all directions except the north-eastern quadrant. Between a direction due west and approximately east-south-east, there are perhaps as many as ten distinct jets. There is in particular a long, strong jet going almost due north which is marked by one of the arrows in Fig. 5.
To assist in understanding the morphology of the region, we have applied a wavelet analysis technique. Extensive literature about wavelet transforms and their application to multiresolution analysis exists (Mallat 1989; Chui 1992; Meyer 1992; Ruskai et al. 1992). Here we use it to emphasise the morphological structure seen on a number of distance scales. We have used the à-trous algorithm with linear scaling functions. The resulting wavelet transforms of the S(1) 1-0 image are shown in Fig. 5. Each scale has double the resolution of the previous one. Scale 1 to 4 thus emphasize structure at a typical resolution of and , respectively. Fig. 5 (bottom left) also shows the composite image that we obtain if all transforms are superposed.
At OMC-1, one arcsecond is subtended by a linear distance of 7 1015 cm. Fig. 5 thus shows respectively linear structures on a scale of 4 1015 cm, 7 1015 cm, 1.5 1016 cm, and structures larger than 3 1016 cm. The jets are best visible at scales 2 and 3. Scale 1 is dominated by small, bright knots which are aligned along the jets. At the largest scale, an image similar to that of Beckwith et al. (1978) is recovered, showing the lobes and emission peaks. In particular, Peak 2 is clearly identified in the south-eastern lobe.
In the wavelet analysis, linear structures show up most clearly in the intermediate scale images, and we take this to be a signature of the jets. They have a typical projected length of cm; a typical width is , which - after correction for the seeing conditions - implies an intrinsic width corresponding to cm. The jets appear to be made up of a string of emission knots, sometimes discernibly connected by weaker, linear emission, which shows up both in the intermediate scale image and at the smaller scale (Fig. 5). The blobs are fairly evenly spaced and with a typical separation of . This projects to a few times 1016 cm. At the end of many of the jets are bright heads, whose S(1) 1-0 luminosity is typically a few times 10-4 .
We may use Fig. 5 to assist in tracing the jets back towards the centre. Allen & Burton (1993) conjectured that the features they saw could have come from IRc2 or BN; from our image, the latter would appear to be ruled out by the line of the northern jet, although the other jets do not discriminate so clearly. In the north-western quadrant and north of east-south-east it is more difficult to trace individual jets through the cloud.
The northern jet passes through a bright ring of emission and emerges much weaker in intensity. If the jet and ring structures are indeed connected, it would appear as if the jet hit a clump of denser material, causing a radially expanding density wave, while passing through.
The jets do not show any measurable broadening in the outward direction. We have estimated opening angles for two of the jets - the northern one and one in the south-east, marked on Fig. 5 (top left) by taking as jet length only the distance over which they can clearly be traced. This provides as upper limits and , respectively. If IRc2 is taken as source, these values are reduced by about a third. The collimation factor is thus of the order of 10.
3.2. Other H2 transitions and continuum emission
In Fig. 6 we show OMC-1 in the light of the H2 transitions S(3) 2-1, S(3) 3-2, O(5) 8-6 and a continuum wavelength. S(3) 2-1 has an excitation temperature of about 14 000 K, i.e. twice as high as S(1) 1-0. The morphology of OMC-1 is nonetheless similar in both lines. Although S(3) 2-1 is considerably weaker, all structures seen in S(1) 1-0 are also present in S(3) 2-1. The situation changes drastically if we turn to the transitions which originate from higher levels. S(3) 3-2 and O(5) 8-6 with excitation temperatures of about 19 000 K and 39 000 K respectively do not show any of the numerous linear structures visible at the lower temperatures. There only remain two larger nebulosities, one around the BN object and one stretching to the north-east of IRc2. The nebulosity near BN seems to have a bipolar structure. Apparently, there are thus two different regimes of H2 emission, a cool one which displays the pattern of numerous linear features and a weaker hot and more compact one which presumably is excited under more extreme shock conditions.
We also have obtained a series of continuum images at various wavelengths. We do not find any significant changes at different wavelengths, so in Fig. 6 we show the sum of the images taken at 2.098 and 2.334 µm. There are a number of stellar sources, many of which have already been catalogued (Lonsdale et al. 1982; McCaughrean & Stauffer 1994). There is weak nebular continuum emission which is most likely due to hydrogen recombination radiation or dust emission. The similarity of the continuum image with the H2 O(5) 8-6 may indicate that hot molecular hydrogen either is located close to the ionisation front or even coexists in protected pockets with ionised hydrogen.
© European Southern Observatory (ESO) 1997
Online publication: July 8, 1998