Astron. Astrophys. 331, 669-696 (1998)

1. Introduction

The link between the structure of interstellar molecular clouds and the star formation process is not yet thoroughly disclosed. Valuable clues, though, exist.

Firstly, the support of interstellar matter against self-gravity is non-thermal and the loss of non-thermal support is therefore one of the prerequisites to star formation. This support is kinetic and magnetic, because on average, the kinetic energy density of non-thermal motions is of the same order as the magnetic pressure, according to measurements of the magnetic field intensity (Myers & Goodman 1988 a, b; Crutcher et al. 1993). It may be magnetohydrodynamic turbulence, or waves, or a combination of both. If the non-thermal support is turbulent, the classical Jeans criterion may be inverted, as proposed by Bonazzola et al. (1987; 1992). The largest scales, predicted to be the more unstable gravitationally when support is only thermal and therefore scale-free, are preferentially stabilized by turbulence. It is the formation of a gradient of turbulent pressure which stabilizes the largest masses (Panis & Pérault 1998). The point of importance is that turbulence and/or waves make the multiscale environment of a gas condensation possibly as important as its temperature and density in determining its gravitational stability (see the reviews of Puget & Falgarone 1990; Fukui & Mizuno, 1991; Myers 1991). This is in opposition to what the Jeans criterion for the growth of gravitational instability stated. Several scenarios have been built to describe the momentum and energy exchange between scales (Henriksen & Turner 1984; Falgarone & Puget 1986; Chièze 1987). They rely on the scaling laws observed between the size and linewidth of clouds, and their mass and size (Larson 1981; Myers 1983; Dame et al. 1986; Solomon et al. 1987).

Secondly, molecular line surveys of interstellar clouds combined with searches for young embedded stellar objects have revealed that stars form in dense cores of several tenths of parsec with considerably less non-thermal support than more dilute regions at the same scale (Beichman et al. 1986; Lada 1992; Phelps & Lada 1997). These cores are bright in the and lines and are often found within massive filaments, at much larger scale (Bally et al. 1987, 1991). A close correlation has been found in nearby clouds between the brightest regions in and the 'cold' fraction of their far-infrared emission at 100 µm, cold here meaning the fraction of the 100µm emission which has a color ratio (60µm)/ (100µm) , the average Galactic value (Laureijs, Clark & Prusti 1991). Regions of low (60µm)/ (100µm) ratio have been shown to be excellent tracers of high column density structures (Abergel et al. 1994). These 'cold' regions, which are often filamentary, are also the large scale structures within which the regions of star formation are embedded. The fact that dense cores have sizes of the order of a few tenths of pc and that the distribution of gas column density at this scale is often consistent with a r^-2 average density distribution (e.g. Zhou et al. 1993) has promoted an attractively simple picture of cold thermally supported self-gravitating structures gravitationally unstable and forming a star after years of self-similar collapse (Shu, Adams & Lizano 1987). This scenario is somewhat invalidated by the existence of clusters of young stars inside of a single core and the detection of small scale structure within dense cores down to about 0.01 pc (Langer et al. 1995; Lemme et al. 1995), suggesting that the initial conditions to the collapse in a dense core may therefore be more complex than those prevailing in an isothermal self-gravitating sphere. These dense cores and massive filaments fill only a tiny fraction of the cloud volume traced by the lines (Ungerechts & Thaddeus 1987; Maddalena et al. 1986). Seen in , molecular clouds appear fractal, down to an observed threshold as small as 1000 AU, and possibly to AU, their average density is low and dense gas fills only an extremely small fraction of their volume (Falgarone, Phillips & Walker 1991; Falgarone, Puget & Pérault 1992; Stutzki 1993; Falgarone & Phillips, 1996). Another prerequisite to star formation is therefore the formation of dense cores and massive filaments of large column density ( ) out of the more porous medium of low column density ( a few ) traced by the lines.

The IRAM key-project of mapping a large fraction of the environment of dense cores which have not yet formed stars, was intended to probe those regions where the dissipation of the non-thermal support is thought to take place, i.e. in the vicinity of dense cores which have not yet formed stars and may therefore be young, or still forming. The originality of the project consisted in mapping a large fraction of the core environment at high angular resolution, to investigate the small scale structure of the transition zone between the ambient molecular medium and dense cores. The second objective of this project is the analysis of the structural properties of dense cores, prior to star formation, speculating that the characteristics of such starless cores illustrate the initial conditions of the star formation process. A third objective is the search for a break in the scaling laws, assigned to turbulence, between size and velocity dispersion of the clouds.

After a presentation of the target fields (Sect. 2) and a summary of the observations (Sect. 3), the essential characteristics of the maps and their spectral properties are presented (Sect. 4). Sect. 5 is devoted to a straightforward interpretation of these properties and to the constraints inferred upon the structure of the CO emitting gas. The statistical analysis of the velocity fields will appear in Pérault et al. (1998). The mass spectra of the observed structures, and the size-linewidth and mass-size scaling laws disclosed among these structures, is presented for one of the fields in Heithausen et al. (1998).

© European Southern Observatory (ESO) 1998

Online publication: February 16, 1998
helpdesk.link@springer.de