2. Observations and reductions
The observations were carried out during the ISO revolutions 271 and 276 (August 13th and 18th 1996). The observed positions are displayed in Fig. 1. The three ON positions (referred to as O1, O2, and O3) were selected to coincide with the IRAS 12, 25, and 100 µm maxima which are at clearly distinct places in the cloud. Two reference positions (R1 and R2) were selected outside the cloud.
The observations were performed in nine filters which are listed in Table 1. Aperture size was , the integration time was 64 sec for the 3.29, 3.6, and 4.85 µm filters and 32 sec for the others. The sparse map Observing Templates (AOTs PHT17/18/19) were used (Klaas et al. 1994). A separate sparse map was performed in each filter. In this observing mode the detector remains switched on during the whole measuring cycle. Since the sky brightness changes only a few per cent between the different ON- and REF-positions in a map this method minimizes drift effects.
Table 1. ISOPHOT filters used in the photometry: central wavelength , width , see Klaas et al. (1994). Internal and external statistical errors, and , are given in percent of the total observed sky brightness.
The first and the last observation of each sparse map were followed by a measurement of the on board calibration source FCS1 (Lemke et al. 1996), which was heated to give a signal corresponding to the expected sky brightness. Presently, the FCS calibration for the large apertures has not been completed. Therefore, in this analysis we have used the zodiacal emission as calibration source. We show in Fig. 2 the zodiacal emission spectrum between 2.2 and 16 µm. Three different measurements were used: (1) ISOPHOT-S observation at , calibrated with standard stars (HR 6705 and HR 6688) to an absolute accuracy of 20% (Ábrahám et al. 1997); (2) ISOCAM-CVF observation at , (Reach et al. 1996); and (3) COBE/DIRBE observation at the south ecliptic pole (Hauser 1996). All spectra were scaled to the position , , applying the well established global brightness distribution of the zodiacal light (see Leinert et al. 1997).
We also show in Fig. 2 the results of our ISOPHOT-P1 photometry at the position R2 near G 300.2-16.8. We have used from each sparse map the last reference position measurement for which the detector drift effects are minimized (see Fig. 3). The default detector responsivity (1.05 A/W) has been adopted for the G 300.2-16.8 data points displayed in Fig. 2. It can be seen that while the shape of our measured ISOPHOT-P1 zodiacal emission spectrum is similar to the DIRBE/ISOCAM-CVF/ISOPHOT-S spectrum there are differences from filter to filter and the absolute level is higher at the longer wavelengths. Based on the comparison with the zodiacal emission spectrum we have thus adopted a responsivity correction for each filter, i.e. we use the reliable zodiacal emission values to calibrate our measurements. This method is further justified by the calibration of the signal dependence on aperture size available so far only for the 11.5 µm filter. Applying the correction factor of 0.55 between stellar calibrations in the 53 aperture and the 180 aperture used here, together with the actual responsivity of 2.16 A/W of the FCS calibration, gives good agreement between both methods.
The data reduction was performed using the ISOPHOT Interactive
Analysis programme (PIA) Version V5.1 (Gabriel et al. 1996). The
following reduction steps were applied:
We display the calibrated measurements for the nine sparse maps in Fig. 3. For the six filter bands between 7.3 and 16.0 µm there is a clear excess signal in the ON positions whereas for the three bands at 3.29, 3.6, and 4.85 µm only an upper limit can be derived. The detector drift has been modelled by fitting through the reference position data points a curve of the form where is the surface brightness and t is the time elapsed since the beginning of the sparse map.
After correction with this fitted line the final spectra of cirrus emission at the three ON positions are obtained and are shown in Fig. 4. Based on this observed energy distribution, modelling of the UIR band emission becomes possible, if gaussian profiles are assumed for the spectral features observed in other galactic emission regions. At the 3.29 µm feature, however, only an upper limit was detected and no separate 6.2 µm band measurement is available. Therefore a model spectrum would not be unique and its presentation is suppressed here.
Statistical errors have been estimated by two methods: (1) internal errors for each measurement are obtained from the PIA analysis; (2) external errors are obtained from comparison of two independent measurements for O1, O2, O3 at = 10, 11.3, 12.8, and 16 µm, and from comparison of all ON and REF point measurements at 3.29, 3.6, and 4.84 µm. At 7.3 and 7.7 µm only one measurement was available for each ON position; upper limits for external errors were estimated from the scatter of these values. Estimates of the internal and external statistical errors are given in Table 1. Relative filter-to-filter calibration errors depend on the accuracy of the zodiacal emission spectral shape and the statistical accuracy of the G 300.2-16.8 R2 measurements used in the calibration (see Fig. 2). We estimate that this accuracy is . The absolute calibration accuracy depends also on the absolute zodiacal emission brightness. Additional information is provided by comparison with IRAS 12 µm emission measured at the positions O1, O2, O3 (see Sect. 3). We estimate that the absolute calibration is accurate to 30%.
© European Southern Observatory (ESO) 1998
Online publication: February 16, 1998