Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 358, 65-71 (2000)

Previous Section Next Section Title Page Table of Contents

2. Observations and data reduction

2.1. Observations

We used the 19-channel bolometer of the MPIfR (Kreysa et al. 1993) in the Nasmyth focus of the IRAM 30-m telescope during the spring session in 1997. The weather conditions were fair during our night-time runs, with rather stable zenith opacities between 0.2 and 0.3 at 230 GHz.

The 19 channels of the bolometer are arranged in a closely packed hexagonal array, with beamsizes of 11" FWHM and spacings of 20". The calibration was performed by mapping Uranus every morning. The pointing and focus checks were made at regular intervals during the observations using nearby quasars (in particular 1308+326). The map was centered on the near-infrared peak at [FORMULA], [FORMULA] (Joy et al. 1987). The data were taken in the standard mapping mode where the field is scanned at a constant speed of 4" per second in azimuth, with subsequent scans being spaced by 4" in elevation. The map size was chosen large enough (320" [FORMULA] 270") to assure a good baseline determination in the presence of extended emission. We thus made sure that the field was mapped sufficiently far out perpendicular to the major axis of M 82, such as to trace the full extent of any dust component in the halo. While scanning the subreflector was wobbled at a 2 Hz rate with an amplitude of 45", which is small enough to ensure a stable beam pattern. Due to the small wobbler throw the OFF position was not necessarily free of emission, but using the algorithm of Emerson et al. (1979) this should have no effect on the resulting map. We also mapped Mars once and the point source 3C 273 twice each in order to obtain a detailed beam pattern. In addition, more maps of Mars and other calibration sources were used to properly check the calibration.

2.2. Standard data reduction

The whole field was covered six times. For each coverage we performed the data reduction separately using NIC (GILDAS software package), starting with the subtraction of a third-order baseline.

The scanning mode described in the previous section leads to maps which contain a superposition of a positive and a negative image of the source (double-beam maps). In principle the deconvolution can be done by dividing the Fourier transform (FT) of the measured intensity distribution by the FT of the wobble function and transforming back to image space. As the FT of the wobble function is a sine wave with zeroes at the origin and at harmonics of the inverse wobbler throw, it would cause problems at these spatial frequencies. To avoid these problems we used the algorithm of Emerson et al. (1979) to restore the double-beam maps into equivalent single-beam maps.

Thereafter we combined the 19 channels and calibrated the resulting map with respect to Mars and Uranus (assuming brightness temperatures of 198.5 K and 97.2 K respectively). In a last step the four best coverages were averaged and regridded onto equatorial (B1950) coordinates. (Two coverages were ignored because of bad S/N ratios.) This final map has a sensitivity of 5 mJy per beam and an angular resolution of 12" (HPBW).

2.3. Influence of the error beam

In order to check the influence of the error beam on the extended emission we applied a CLEAN algorithm developed for single-dish maps (Klein & Mack 1995) to each coverage using a radially symmetric antenna pattern. This pattern was obtained in the following way: A field of 4:05 [FORMULA] 3:05 size was mapped centred on 3C 273. This antenna pattern was then rotated in small steps ([FORMULA]), and these individual (rotated) maps were averaged to yield a nearly radially symmetric pattern. The subtraction of the CLEANed map from the original one showed that the error beam contributes less than 25% of the rms noise level to the extended emission ([FORMULA]" away from the central position).

Even the sidelobes can be ignored in our case, because at [FORMULA] 1.2 mm they are of the order of 2%, corresponding to roughly 5 mJy/beam in the M82 maps. Averaging four coverages with different parallactic angles (and accordingly different positions of the sidelobes) the effect of the antenna pattern on the extended emission falls well below the rms noise of our final map.

This extra analysis makes sure that the extended emission seen in our map is real and not affected by the error beam.

2.4. Correction for CO emission

Using the 19-channel bolometer, with its central frequency of 240 GHz and its effective bandwidth of [FORMULA] 80 GHz, the 12CO(2-1) line at 230 GHz might noticeably contribute to the total intensity in the bandpass. To get a reliable information on the distribution of the cold dust, we applied the following correction to our continuum map:

We used a [FORMULA] map of M 82 made with the same telescope in the 12CO(2-1) line (obtained to provide the zero spacings for an investigation with the Plateau de Bure interferometer, Weiß et al., in prep.). This map was measured in the on-the-fly mode. The on-the-fly technique ensures a uniform sampling of the emission over the whole field with good signal-to-noise ratio. Since we knew from our continuum map that the effects of the error beam can be neglected at a frequency of 240 GHz, we did not apply any error beam correction to the CO map.

Integrating over the effective bandwidth of the bolometer, the total continuum flux in the beam at the peak position is given by


where [FORMULA] mJy is the peak flux at 240 GHz.

The integrated intensity of the 12CO(2-1) line at the same position [FORMULA] was taken from our CO map. Expressing this in terms of frequency instead of velocity, the integrated intensity reads [FORMULA].

To obtain the total flux in the 12CO(2-1) line we used the Rayleigh-Jeans Approximation and assumed that [FORMULA]:


With the beam solid angle [FORMULA]sr and the integrated 12CO(2-1) intensity from above we get [FORMULA]. Thus, [FORMULA] at the position of the continuum peak.

If we assume that this value holds for the whole galaxy, which of course is a simplification, we can obtain a CO-corrected continuum map by subtracting the appropriately scaled CO map from our continuum map. The uncorrected and the corrected map are shown in Fig. 1.

[FIGURE] Fig. 1. Maps of M 82 at [FORMULA]1.2 mm. The upper map shows the continuum emission after standard data reduction. The lower one shows the continuum emission, with the CO line emission subtracted. The contours are the same in both images starting at the 2[FORMULA] (10 mJy/beam) level. The beam size of the IRAM 30-m telescope is shown in the lower left corner of each image.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: June 26, 2000