Astron. Astrophys. 363, 1123-1133 (2000)

2. Observations and data reduction

We observed M31 in a wavelength range of with the Long-Wavelength Spectrometer (LWS; Clegg et al. 1996) on board the Infrared Space Observatory (ISO; Kessler et al. 1996). The observations comprised three one-dimensional raster scans with the Target Dedicated Time (TDT) numbers 58002001, 58302603, and 60202005. The combination of these raster scans results in a strip map of 31 positions along the major axis (the position angle is ) of the galaxy centered on the galactic nucleus at and . The observed positions were spaced at , with offsets of relative to the galactic nucleus ( in TDT 58302603, in TDT 58002001, and in TDT 60202005; x is positive in the northeast).

We adopted the Astronomical Observation Template (AOT) LWS01 mode: grating spectroscopy with a medium-resolution of at and at . The observations at each of the raster position consisted of 4 grating scans, with 1.6 seconds of total integration time for each grating position. The grating positions were spaced at 1/2 of the spectral resolution.

The beam size of the LWS derived from observations of Mars was and in FWHM averaged over position angle for the detectors LW1 (around ) and LW4 (around ), respectively (Lloyd 1999). The equivalent disks had nearly equal solid angles for the detectors LW1 and LW4: and , respectively (Lloyd 1999).

We subtracted the detector dark currents from the Standard Processed Data (SPD) of Off-Line Processing (OLP) version 7.0 products, using the LWS Interactive Analysis ¹ (LIA) version 7.2a and 7.3. The ISO Spectral Analysis Package ² (ISAP) version 1.6a was used for the data reduction afterwards. The data affected by cosmic-ray hits were manually removed, when they had not been discarded automatically in the SPD. Then the dark current level of LW1 was reevaluated so that the spectrum is consistent between LW1 and LW2 at the overlapping wavelength region of the two detectors because LW1 has problems with dark-current measurements. We derived the [CII] flux from a single-Gaussian fit with a polynomial baseline of the first or second order. The width, height, and central velocity of the line profile were determined by the fit for a spectrum with a [CII] flux of . On the other hand, only the height and central velocity were determined by the fit for a spectrum; the width for the fit was fixed to that of the instrumental profile for such a spectrum with a limited signal-to-noise ratio. The continuum flux density, , at was derived from the convolution of the LWS spectrum and the IRAS band transmission.

The flux calibration of the LWS grating mode was based on observations of Uranus (Swinyard et al. 1998) for point-like sources. This calibration is accurate to 15% when uncertainty concerned with extended-source correction and dark-current subtraction are excluded (Swinyard et al. 1998). In case of the continuum emission from faint (less than 100 Jy at ) sources such as the central region of M31, the calibration uncertainty can be larger because the instability of detector dark-current often determines the calibration accuracy. This instability affects the calibration more significantly for raster scans than for single-point observations, since the instability is difficult to distinguish from spatial variation in source brightness. Thus, we evaluated the uncertainty in on the basis of the present observations, by comparing the LWS data with the corresponding IRAS data (Sect. 3.2).

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk@link.springer.de