This project is devoted to the analysis of the processes which drive the dissipation of the supersonic non-thermal support of molecular clouds, known to be at the origin of their gravitational stability. One of the outcomes of this energy dissipation is expected to be the formation of thermally supported structures. The formation of dense cores, observed in molecular clouds with an internal velocity dispersion close to thermal, and often associated with extremely young stellar sources, is possibly triggered by the local dissipation of supersonic turbulence.
We have therefore mapped several fields of nearby clouds (d<150pc)
which all contain a starless
dense core of small internal velocity dispersion. The maps include
the core (of size 
0.1 pc) or a fraction of it, and extend over a
large area (several arc minutes, or several tenths of pc) of their
environment, characterized by a non-thermal velocity dispersion.
Maps have been completed in five transitions, 
J=1-0 and J=2-1,
J=1-0 and J=2-1 and 
J=1-0, at high angular resolution
(13'' and 24'' at high and low frequency respectively, with a sampling
of 7.5'' which is Nyquist sampling for the low frequency maps)
and a velocity resolution of 0.05 km s 
.
The spatial resolution of the high frequency maps is therefore
1000 AU.
Between December 1992 and March 1994, 624 hours were scheduled for this
project. It was also scheduled as a backup project of the 345GHz
receiver run of January 1994 (237 hours) but the efficiency of this period
was quite low since only 39.7 hours were used.
A total of 338 hours were spent in effective observations.
322 hours were lost for bad weather or zenith optical depth at 230GHz
larger than 0.4, and technical reasons.
But 92% of this time was lost for bad weather conditions: it means
that the weather at Pico veleta degrades very fast and that the limit
set on 
(230 GHz) was not a real constraint
(either (230 GHz)<0.4 or (230 GHz)>>0.4).
On the average, the time spent
on the telescope has been twice the observing time, the additional overhead
within this observing time
due to calibration and tuning of the receivers being 30%.
The data base consists in 2.7 
spectra.
The final baseline rms noise level ranges between 0.6 and 1.2 K for
the spectra, 0.5 and 0.9 K for the \
spectra, 0.3 and 0.5K for the J=1-0 spectra
and 0.3 and 1K for the spectra.
The data set, because of its size, and the multiplicity of the lines observed, provides several new results, as follows:
(1) despite their low average column density at the parsec scale
( 
), all the
fields observed appear highly structured in all the lines, including
those of .
Unresolved structure is still present in all the
fields and all the lines, in the dense cores and in their lower column
density environment.
It corresponds to size scales < 0.008 pc,
(2) the texture and velocity dispersion of the gas bright in \
and weak in and are significatively different
from those of the gas bright
in or . The former exhibits
filamentary structure with, in some cases, unresolved transverse
dimensions, and aspect ratios 
. Its velocity dispersion
is also much larger than that of the latter,
(3) the uniformity of the excitation temperature ratio of the two
lowest CO rotational transitions
in the three fields, from
the brightest to the weakest detected lines, across the whole
profiles and for both and isotopes, is remarkable,
(4) the optical depths of the and lines reach very large
values, though most of the line profiles are
neither flat-topped nor self-reversed.
We deduce from
these well-defined properties that the uniformity of the CO
line excitation is a very robust property of non star-forming
molecular clouds. The large range of
optical depths and line temperatures over which it is met,
imposes a line formation
mechanism confined to regions smaller than a few 100 AU and denser than
for a kinetic temperature T=20K, yet
denser in colder gas. Furthermore, the line shapes impose
a weak level of radiative coupling among these regions, a
condition met in macroturbulence.
It is interesting to stress the outcomes of this program which are specific to its key-project status:
- on scientific grounds
The large size of the maps and the good signal-to-noise
level in have allowed
the discovery of a filamentary structure, hard to
detect because it is
small-scale structure in almost
transparent gas. Furthermore such structures
are simultaneously small (narrow width) and large scale
(long length) and cannot be recognized in either small maps or
large scale maps at low angular resolution.
It is only the systematic mapping of low brightness and large
areas which allowed the detection of such filamentary structures which
are thought to play a role in the dense core formation.
Statistical work on the distribution of line centroid velocities and the power law relations is under progress. The statitical noise is considerably reduced due to the large size of the maps.
The signal to noise achieved was a compromise between the dynamic range of the maps and the total time devoted to the project. It is good enough to allow some work to be done on the line shapes and line ratios of individual spectra.
- on instrumental grounds
A long term monitoring of the time-dependent variations of the line calibration at the telescope has been achieved. We have taken advantage of the various time scales over which the monitoring was performed to correct for the non-thermal contribution to noise in the line calibration and our method reduces the line calibration error by a factor 1.5 at 2.6 mm and a factor 2 at 1.3 mm.
Six hardware and software problems have been discovered, in particular in the frequency settings, and have now been corrected.
Edith FALGARONE