A
spectrophotometric map in the far-infrared (FIR) of the
was obtained with the
Long-Wavelength Spectrometer (LWS; 43-197
m,
-330)
on board the Infrared Space Observatory (ISO) on October 21, 1996.
The ISO-project is described by Kessler et al. (1996) and
the LWS is described by Clegg et al. (1996) and Swinyard et al. (1996).
The formal map centre, viz.
= 18
29
50
29 and
= 1
15
18
6,
epoch J 2000, coincides to within 10
with the position of the
sub-millimetre source SMM 1. The pointing accuracy in the map is determined as 1
(rms).
The spacings between positions in the map, oriented along the equatorial coordinates, are
100
.
The size of the LWS-map is 8
8
,
corresponding to
D3102 pc2, where D310 denotes the distance
to the
in units of 310 pc.
At each map-point the grating of the LWS was scanned 6 times in fast mode, oversampling the spectral resolution at twice the Nyquist rate. Each position was observed for nearly 15 min. The centre position was re-observed half a year later on April 15, 1997, for a considerably longer integration time (24 spectral scans).
In addition, imaging spectrophotometric data obtained with the Continuous Variable Filter
(CVF; 5-16.5 m,
pxl, pixel-FOV = 6
,
35) of ISO-CAM (Cesarsky et al. 1996) were also analysed. The detailed observing log is given in our
previous paper (Larsson et al. 2000).
The LWS data reductions were done with the interactive analysis package LIA. This pipeline processing was done in two ways. For the long wave detectors LW 1-LW 5, the Relative Spectral Response Functions (RSRFs) of the most recent version (OLP 10) provided an overall better agreement than the older OLP 8 for the detector inter-calibration. However, the OLP 10 RSRFs also contained a number of strong features, which resulted in spurious "lines'' in the reduced spectra (Appendix A). We decided therefore to use the RSRFs of OLP 8, which did not show this behaviour, but scaled to the OLP 10 absolute levels. For the short-wave detectors SW 1-SW 5, no such difference between OLP 8 and OLP 10 was apparent and consequently we used the latest version. A detailed account for the first steps in the LWS reduction is given in Appendix A.
Subsequent post-pipeline processing used the package ISAP.
At each position, the individual LWS scans for each of the 10 detectors were
examined, "deglitched'' and averaged. Corrections were applied to the
"fringed'' spectra in the map. The fringing indicates that the emission is
extended and/or that point sources were not on the optical axis of the LWS, i.e.
the radial distance from the optical axis was typically larger than about 25
.
The absolute flux calibration is better than about 30% (Swinyard et al. 1996).
Overlapping spectral regions of adjacent detectors were generally within 10%
("detector stitching'' uncertainty). At 60 m and 100
m, the correspondance with
broad-band IRAS data is even better than 10% (Larsson et al. 2000).
The ISO-CAM data were reduced with the CIA programs using OLP 8.4. This included dark current
correction, transient correction and deglitching and flat field corrections. The calibrations
were based on OLP 5.4. The proper alignment of the CVF fields was achieved with the aid of
point sources identified in both CVF and broad-band (7 m and 14
m) CAM observations
(Fig. 1).
![]() |
Figure 1:
The from CVF frames synthesised images at 7 ![]() ![]() |
Copyright ESO 2002