Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 336, 991-1006 (1998)

Previous Section Next Section Title Page Table of Contents

2. Observational procedures

2.1. Centimeter observations

The observations of the [FORMULA] @ and (2,2) lines were conducted with the 100-m telescope in December 1993, March and April 1994, and March and April 1995. At the line frequencies of 23.694 GHz and 23.722 GHz the telescope has a full width half-power beamwidth (FWHP) of [FORMULA], a beam efficiency of [FORMULA] and an antenna efficiency of [FORMULA]. The receiver was a K-band maser with a system temperature (including atmospheric contribution) of normally less than 60 K and never more than 70 K. The (1,1) and (2,2) inversion lines of [FORMULA] are separated by 28 MHz and thus could be measured simultaneously. The spectrometer was a 1024 channel autocorrelator which we operated as two 512 channel receivers, each with a bandwidth of 12.5 MHz for each line. The resulting channel spacing was 0.3 km s-1 . We observed a Nyquist sampled map with [FORMULA] spacing in the position switching mode, using a reference position [FORMULA] west of the source. Usually three on-source scans of 60 seconds duration each were made for one reference spectrum. Every position was observed twice, on separate days. Positions toward W 3 Main and the W 3(OH) region were observed up to 4 times to improve the signal-to- noise ratio. Thus, we obtained a total of 4800 scans for the three observing runs. Since the [FORMULA] @ and (2,2) inversion lines have been observed simultaneously, calibration or pointing differences can be ruled out when comparing emission.

During our last run in March 1995, we observed 28 positions toward W 3 Main in the (J,K) = (3,3) inversion line of [FORMULA] at a line frequency of 23.870 GHz. We obtained a Nyquist sampled map in [FORMULA] @ for the central region of W 3 Main with the same conditions given above.

During all of our observations, we checked our pointing by continuum scans through W 3(OH) and NGC7027 prior to the line observations. The rms pointing was accurate to [FORMULA] [FORMULA]. The pointing scans also allowed us to obtain an estimate for the flux of the continuum sources and to calibrate the line data. The line temperatures are then given on a main beam brightness scale, [FORMULA], corrected for the atmospheric attenuation and the variation of telescope gain with elevation.

2.2. Near-infrared observations

The near-infrared observations presented here were made at the f/10 focus of the University of Hawaii (UH) 88-inch Telescope on Mauna Kea by J. T. Rayner during September 1990, using the UH Near Infrared Camera and Multiple Object Spectrometer (NICMOS) in its [FORMULA]/pixel "mapping" mode (Hodapp et al. 1992). The camera incorporated the 256[FORMULA]256 NICMOS-3 HgCdTe detector array sensitive from 1-2.5µm, produced by the Rockwell International Science Center, under contract from the University of Arizona, for the NICMOS Project (Thompson et al. 1989). A [FORMULA] filter (2.1 µm, FWHM [FORMULA]m) was used in preference to the standard K filter because then there is a reduction in thermal background from a 0.1 µm shift of the central wavelength to shorter wavelengths (Wainscoat & Cowie 1991).

Three [FORMULA][FORMULA][FORMULA] mosaics were obtained along the line connecting W 3 Main to W 3(OH); centered on W 3(OH), the O8 star BD[FORMULA] 411 NW of W 3(OH), and W 3 IRS5. Each mosaic consists of 27 images covering a 3[FORMULA]3 position grid with successive position offset by the field of view. Three 30 second dithered images were taken at each position before offsetting the telescope about [FORMULA] to a region free from detectable nebulosity and bright stars, where three dithered "sky" images were taken. The three mosaics have been assembled into the final mosaic shown in Plate I.

[FIGURE] Plate I shows the mosaic of 81 images obtained in the [FORMULA]-band (2.1 µm) along the high density layer from W 3 Main to W 3 OH. The mosaic consists of three square sub-mosaics centered on W 3(OH), the O8 star BD[FORMULA] 411 NW of W 3(OH), and W 3 IRS5. The bright nebulae in the mosaic are found toward embedded compact and ultracompact H II regions. By visual inspection, we have identified five distinct stellar clusters in the mosaic.

From the astrometry of visible stars in the image, the measured pixel scale is 0:00755[FORMULA]0:00002/pixel. By scanning a star across the array we have determined that the difference between the point-source flat-field and the extended dome flat-field is better than 1%. IRAF was used to process the raw images. A median-filtered sky image was constructed by taking the true median of a stack containing unregistered images taken over a period of about 15 minutes. Typically 10 images of regions containing no nebulosity were required for a good median filtered sky image (photometry better then 0.5%), a procedure which worked well even on relatively crowded fields. These images were first flat-fielded by using a dome flat-field (incandescent lights on minus lights off). The mosaics (i.e. co-added images) were assembled by adding images into a template image at positions determined by calculating offsets from stars common to more than one image. Due to the crowded nature of the sources and the large pixel size, the [FORMULA]-band mosaic is not suitable for photometry of the embedded stellar clusters. It has been used to select regions for high-resolution infrared imaging, and in this paper to compare with the large scale distribution of ammonia in the W 3 HDL. On isolated point sources the image reaches a 3[FORMULA] limit of [FORMULA]=17.5 magnitudes. A fits version of the full [FORMULA]-band mosaic is available by anonymous ftp to hubble.ifa.hawaii.edu; the file is W 3.big.fits.gz and is found in the directory pub/rayner/W 3.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: July 27, 1998