transitions
The CO and 
observations of the transitions
at 115 and 110 GHz, respectively, have been carried out in March and April
1991 with the Center for Astrophysics (CfA) 1.2 m telescope in Cambridge,
Massachusetts. The full beam size at half maximum is 8.7' at 115 GHz. 1407
CO spectra, spaced every 
along a regular (l, b) coordinate grid,
cover a total of 
. A set of positions has been
subsequently selected to sample the cloud in , covering the
whole range of observed peak temperatures and line velocities (see
Fig. 1 (click here)).
The receiver is a sensitive liquid helium-cooled SIS mixer, with a typical
single-sideband noise temperature of about 65 K at the time of the
observations. Spectral resolution is provided by a 256-channel filter-bank
spectrometer. We used a frequency resolution of 250 kHz, i.e. a velocity
resolution of 0.65 and 0.68 km/s at 115 and 110 GHz, respectively, centered
on 0 km/s in the local standard of rest. Because of telluric CO emission at
velocities close to the range of interest and because of the large width of
the expected lines, all spectra have been observed in position-switching
mode using two reference positions to straddle the source in elevation. The
latter positions have been checked by frequency switching to be free of
appreciable CO emission. Position switching was also preferred because it
yields very flat spectral baselines that generally allow a clear
decomposition of the multiple lines into gaussian components. Only linear
baselines have been removed from the raw spectra. For overall system
temperatures including the atmosphere of 
in CO and
in , typical integration times of at most a
few minutes were needed to achieve a rms noise of 0.1 K in radiation
temperature.
Figure 1: Map of the CO emission at 115 GHz, integrated between -10 and
+10 km/s in radial velocity, from the eastern cloud of the Cepheus
complex. The angular resolution is 8.7'. Contours increase from 1.3 K
km/s (
) with a step of 2 K km/s. The positions sampled at 110,
220 GHz (x) and 230 GHz (
) have been superimposed
Calibration and correction for atmospheric absorption have been made by briefly rotating a room-temperature blackbody wheel in front of the feed horn before each scan. The temperature and opacity of atmospheric water vapor have been estimated by antenna tipping as often as weather variations required. To check the pointing and calibration, a reference position has been repeatedly observed in Cepheus; day-to-day intensity variations amount to 4%. The resulting radiation temperatures have been scaled by a factor of 1.22 to correct them to absolute brightness temperatures as recommended by Bronfman et al. (1988) after their re-evaluation of the beam efficiency.
transitions
The higher-frequency transition of CO and have been observed
during the winters of 1994 and 1995 with the POM-2 2.5 m telescope on the
Plateau de Bure in France. The beam size at half intensity of the antenna is
2'.3 at 230 GHz. The telescope is equipped with an SIS receiver, cooled by
a 2 K cryogenerator, and a 256-channel autocorrelator back-end. The CO and
data, at 230 and 220 GHz, respectively, have been taken with a
312 kHz resolution that corresponds to velocity resolutions of 0.41 and 0.43
km/s. As for the CfA telescope, the back-end was centered on 0 km/s in the
local standard of rest. In order to reduce the integration time, and because
the telluric CO line lay more than 5 km/s away from those of interest,
frequency switching has been adopted, with a switching interval of 18 MHz.
To allow a comparison of the 
and 
spectra,
the POM-2 data have been smoothed to the CfA beam resolution. To do so, each
selected position observed in has been mapped in
; the map, centered on the position of the
spectrum, consists of 4 by 4 positions spaced every 2' along a regular
(l, b) coordinate grid. After folding the raw spectra, each map has been
convolved with the CfA antenna pattern. To achieve an rms noise of 0.05 K on
the synthesized spectra, individual scans have been integrated to a noise
level lower than 0.4 K. For system temperatures of 500 to 1500 K for both
lines, typical integrations were 2 min per spectrum. As a final step,
baselines of order 3 to 5 have been subtracted from the synthesized spectra.
The POM-2 radiation temperature scale has been corrected for the main beam
efficiency and for atmospheric absorption by means of ambient hot-cold
calibration blackbodies at the telescope and a standard atmospheric model
(Kutner 1978). Pointing and calibration have been tested in
Orion at the beginning of each observing run and on the reference position
in Cepheus every hour. Observed intensity variations were no more than 7%.
The radiation temperatures have been scaled by a factor of 1.3 to match
the expected intensity from the Orion reference position. The calibration
of POM-2 has been checked against the AT&T Bell Laboratories 7 m telescope
using the Barnard 5 source to study multi CO transitions in Orion A (Castets
et al. 1990).