All Figures

$\begin{figure} \par\includegraphics[scale=0.45,angle=0.,clip]{h4471f1.eps} \end{figure}$ Figure 1: A schematic representation of the P32 observing technique. The sky is sampled using a combination of a coarse spacecraft raster in the satellite Y and Z coordinates (horizontal and vertical directions on the figure) and finely sampled sweeps of the focal plane chopper in the Y coordinate. The spacecraft raster (in this example a $4\times 3$ raster) starts at the maximum value of Z and the minimum value of Y and proceeds as shown by the arrows. The position angle $\beta$ of the first scan leg in Y according to normal astronomical convention (taken positive E from N) is related to the satellite roll angle $\alpha$ by $\beta = \alpha +90^\circ$ . At each spacecraft pointing direction a chopper sweep of length 180 arcsec (the diameter of the unvignetted field of view of ISO) is made with a sampling of 1/3 of the detector pixel separation. This results in chopper sweeps of 13 positions separated by 15 arcsec for observations using the C100 detector array (as depicted in this example), and chopper sweeps of 7 positions sampled at 31 arcsec for the C200 detector array. The spacecraft raster sampling interval in Y is set to be a multiple of the chopper sampling interval (in the example depicted it is set to 6 chopper sampling intervals). A series of uniformly sampled linear scans for each detector pixel in Y is obtained in this fashion, built up from a sequence of registered and overlapping chopper sweeps obtained on each spacecraft scan leg in Y. The sampling of the target field in Z is controlled entirely by the spacecraft raster sampling interval in Z. In the example depicted the C100 detector detector array is stepped in intervals of 1.5 times the detector pixel separation in Z. For this particular case, the linear scans from different C100 detector pixels can be combined to sample the sky at half the detector pixel separation in Z, resulting in an overall sampling of $15\times 23$ arcsec over the target field.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.55,angle=90.,clip]{h4471f2.eps} \end{figure}$ Figure 2: Pointing directions observed towards Ceres at 105 $~{\mu}$ m. The spacecraft raster dimension was $3 \times 3$ in the spacecraft coordinates $Y \times Z$ . Sky sampling in Y is achieved through a combination of chopper scans in Y and repointings of the spacecraft. In Z the sky sampling is determined alone by the spacecraft repointing interval. Each pointing direction for 11 833 individual data samples for the central pixel of the C100 detector array is plotted as a cross. The rectangular grid is the "P32 natural grid'' (see text) for this observation, on which the sky brightness distribution is to be solved. The grid sampling is $14.99\times 45.98$ arcsec. Only points with the on target flag set (i.e. notincluding slews) are plotted. The pointings typically lie within1 arcsec of the centre of each pixel of the P32 natural grid.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.52,clip]{h4471f3.eps} \end{figure}$ Figure 3: Example of the transient response of a Ge:Ga detector following an upwards step to a constant level of illumination at 100 $~\rm\mu m$ . The hook response is shown in the inset.
Open with DEXTER

In the text

$\begin{figure} \par\epsfig{file=h4471f4.eps, width=8cm,clip=} \end{figure}$ Figure 4: A schematic representation of the response of the semi-empirical detector model used in the P32 algorithm to an upwards step in illumination from $S_{\rm\infty p}$ to $S_{\rm\infty }$ . The overall response S is the sum of a "slow'' signal response S₁ and a "fast'' signal response S₂ (see Sect. 3.1). $S_{\rm 1p}$ and $S_{\rm 2p}$ respectively denote the signal response of these two components immediately prior to the step in illumination.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.52,clip]{h4471f5.eps} \end{figure}$ Figure 5: Example of a model fit (curve) to the response of the central pixel of the C100 array to a constant FCS illumination starting at t=0 s. Prior to the FCS illumination the detector was viewing blank sky. The horizontal line indicates the fitted value of the illumination. No data was recorded in the first two seconds following FCS switch-on.
Open with DEXTER

In the text

$\begin{figure} \par\mbox{\subfigure[]{ \includegraphics[scale=0.67,clip]{h4471f6... ...} \subfigure[]{ \includegraphics[scale=0.67,clip]{h4471f6d.eps} }}\end{figure}$ Figure 6: a) Values of the parameter $\beta _{\rm 1}$ for pixel 8 of the C100 array found from fits to the detector response to FCS exposures, plotted versus FCS illumination. The solid line is a fit of the function $\beta _{\rm 1}=\beta _{\rm 10}+\beta _{\rm 11}S_{\rm\infty }^{\beta _{\rm 12}}$ with $\beta _{\rm 10}=0.96$ , $\beta _{\rm 11}=-0.28$ , and $\beta _{\rm 12}=0.075$ . These are the parameters adopted for the model giving the default illumination dependence of $\beta _{\rm 1}$ for this pixel. b) As a), but for the parameter $\tau _{\rm 1}$ (in seconds). The solid line is a fit of the function $\tau _{\rm 1}=\tau _{\rm 10}+\tau _{\rm 11}S_{\rm\infty }^{-\tau _{\rm 12}}$ with $\tau _{\rm 10}=7.73$ s, $\tau _{\rm 11}=11.60$ , and $\tau _{\rm 12}=-1.28$ . These are the parameters adopted for the model giving the default illumination dependence of $\tau _{\rm 1}$ for this pixel. c) Values of the parameter $\beta _{\rm 2}$ for the central pixel of the C100 array found from self calibration optimisation of the model on observations of standard point source calibrators, plotted versus illumination. The solid line is a fit of the function $\beta _{\rm 2}=\beta _{\rm 20}+\beta _{\rm 21}S_{\rm illum}^{\beta _{\rm 22}}$ with $\beta _{\rm 20}=1.171$ , $\beta _{\rm 21}=-0.870$ , and $\beta _{\rm 22}=-0.0145$ , where $S_{\rm illum}$ is the known illumination of the detector pixel due to the point source calibrator. These are the parameters adopted for the model giving the default illumination dependence of $\beta _{\rm 2}$ for this pixel. d) As c), but for the parameter $\tau _{\rm 2}$ (in seconds). The solid line is a fit of the function $\tau _{\rm 2}=\tau _{\rm 20}+\tau _{\rm 21}S_{\rm illum}^{-\tau _{\rm 22}}$ with $\tau_{\rm 20}=0.333~$ s, $\tau _{\rm 21}=0.381$ , and $\tau _{\rm 22}=0.584$ . These are the parameters adopted for the model giving the default illumination dependence of $\tau _{\rm 2}$ for this pixel.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.8,clip]{h4471f7.eps} \end{figure}$ Figure 7: Iterative scheme for transient correction (see Sect. 4.2).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=15.0cm,clip]{h4471f8.eps} \end{figure}$ Figure 8: Detail from a signal timeline from an observation of the standard calibrator Ceres, observed in the C105 filter. The green line shows the fit to the observed signal (shown in red) from pixel 1 of the C100 array, as found by the P32 algorithm for correction of the transient response behaviour of this pixel. 5 chopper sweeps, each comprising 13 pointing directions ("chopper plateaus'') are shown, divided by vertical dotted lines separated in time by a duration of 6.1 s for each sweep. The upper line in the figure shows the fine pointing flag; whereas the first chopper sweep shown was made while the spacecraft was slewing between fine pointings (flag value 0), the subsequent chopper sweeps were made on a single fine pointing (flag value 1). The remaining horizontal lines denote masks at the full time resolution of the data denoting glitches, readout status and whether or not a solution for the illumination could be found by the algorithm. The overall solution for the sky illumination for this detector pixel is given by the histogram in black. As described in the text, this overall solution is calculated from a combination of individual solutions for illumination on each successive chopper plateau, which are shown here as purple bars. For the very brightest sources, as shown in this example, the derived illuminations can be up to a factor of 6 brighter than the raw data for pixels in the C100 array. In this example the hook response is well seen in each chopper sweep at the 9th and 10th chopper plateaus.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=15.0cm,clip]{h4471f9.eps} \end{figure}$ Figure 9: Detail from a timeline of signal versus time from another observation of Ceres, this time observed in the C160 filter using the C200 detector array. The key to the plot is as given in the caption to Fig. 8. This example shows the transition from a spacecraft fine pointing for which the chopper is sweeping through beam sidelobes to a fine pointing where the chopper samples the beam kernel. In this case the red line corresponding to the data is not seen during the intervening spacecraft slew, as it is exactly overplotted by the green line showing the fit to the data (see text). The chopper sweeps on the second fine pointing encompassing the beam kernel show a typical enhancement of the source-background contrast induced by the algorithm for correction of the transient response.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.6,clip]{h4471f10.eps} \end{figure}$ Figure 10: The measured integrated and colour-corrected flux densities of the standard calibrators Ceres, Vesta and Neptune plotted against the nominal flux densities. The observations were done in various filters using the C100 detector. The photometry derived with and without correction for the transient response behaviour is shown with crosses and diamonds, respectively. The solid line shows the expected trend for equality between the measured and nominal flux densities.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.5,clip]{h4471f11.eps} \end{figure}$ Figure 11: The measured integrated flux densities of the faint standard star HR 1654 plotted against the predicted flux densities from a stellar model. The observations were done in various filters using the C100 detector. The photometry derived with and without correction for the transient response behaviour is shown with crosses and stars, respectively. The solid line shows the expected trend for equality between the measured flux densities and those predicted from the stellar model.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.5,clip]{h4471f12.eps} \end{figure}$ Figure 12: Integrated and colour-corrected flux densities of Virgo cluster galaxies measured in the ISO C100 filter versus the corresponding flux densities measured by IRAS in its 100 $~\rm\mu m$ band (taken from Fig. 7 of Tuffs et al. 2002a). The dotted line represents the relation $\rm ISO/IRAS=0.82$ . The star symbols are used for galaxies not fully covered by the spacecraft raster.
Open with DEXTER

In the text

$\begin{figure} \par\epsfig{file=h4471f13.eps, width=8.0cm,clip=} \end{figure}$ Figure 13: Brightness profiles (in MJy/str) along the Y spacecraftdirection through the standard star HR 1654 at 100 $~\rm\mu m$ . The solid line represents the brightness profile obtained after processing with the P32 algorithm, while the dotted line shows the profile derived from identically processed data, but without correction for the transientresponse of the detector.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.72,clip=]{h4471f14a.eps}\par\includegraphics[scale=0.72,clip=]{h4471f14b.eps} \end{figure}$ Figure 14: Top: contour map of Ceres in the C105 filter after responsivity drift correction. The contour levels are 5 . 2ⁿ MJy/sr for all integers n with $1 \le n \le 10$ . The measured integrated flux density after background subtraction is 98.4 Jy $\pm~0.2$ Jy (random) $\pm$ 6% (systematic). The actual flux density of this standard calibrator (from a stellar model) is 109 Jy. As in Fig. 2, the the map pixels of the P32 natural grid have been overlaid. Bottom: contour map of Ceres in the C105 filter without correction for the transient response behaviour of the detector. The contour levels are 5 . 2ⁿ MJy/sr for all integers n with $1~\le~n~\le~7$ . The The measured integrated flux density after background subtraction is 28.0 Jy $\pm~0.07$ Jy (random) $\pm$ $4\%$ (systematic).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.72]{h4471f15a.eps}\par\includegraphics[scale=0.72]{h4471f15b.eps} \end{figure}$ Figure 15: The interacting galaxy pair KPG 347 observed in the C160filter. Top: after processing with the P32 algorithm. Bottom: with identical processing, except that the correction for the transient response behaviour of the detector has been omitted. The map pixels corresponding to the P32 natural grid (see Sect. 2.1) have been overlaid. In both maps, contours are logarithmic, at levels 1, 2, 4, 8, 16, 32, 64 MJy/sr. The peak brightnesses are 126 and 109 MJy/str for the maps with and without processing with the P32 algorithm, respectively.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.65,clip]{h4471f16.eps} \end{figure}$ Figure 16: A $27\times 27$ arcmin field containing the galaxy M 101, as mapped in the C100 filter, after processing with the P32 algorithm. The map sampling is $15\times 23$ arcsec.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[scale=0.52,clip]{h4471f3.eps} \end{figure}$	Figure 3: Example of the transient response of a Ge:Ga detector following an upwards step to a constant level of illumination at 100 $~\rm\mu m$ . The hook response is shown in the inset.
Open with DEXTER

$\begin{figure} \par\includegraphics[scale=0.52,clip]{h4471f5.eps} \end{figure}$	Figure 5: Example of a model fit (curve) to the response of the central pixel of the C100 array to a constant FCS illumination starting at t=0 s. Prior to the FCS illumination the detector was viewing blank sky. The horizontal line indicates the fitted value of the illumination. No data was recorded in the first two seconds following FCS switch-on.
Open with DEXTER

$\begin{figure} \par\includegraphics[scale=0.8,clip]{h4471f7.eps} \end{figure}$	Figure 7: Iterative scheme for transient correction (see Sect. 4.2).
Open with DEXTER

$\begin{figure} \par\includegraphics[scale=0.5,clip]{h4471f12.eps} \end{figure}$	Figure 12: Integrated and colour-corrected flux densities of Virgo cluster galaxies measured in the ISO C100 filter versus the corresponding flux densities measured by IRAS in its 100 $~\rm\mu m$ band (taken from Fig. 7 of Tuffs et al. 2002a). The dotted line represents the relation $\rm ISO/IRAS=0.82$ . The star symbols are used for galaxies not fully covered by the spacecraft raster.
Open with DEXTER

$\begin{figure} \par\epsfig{file=h4471f13.eps, width=8.0cm,clip=} \end{figure}$	Figure 13: Brightness profiles (in MJy/str) along the Y spacecraftdirection through the standard star HR 1654 at 100 $~\rm\mu m$ . The solid line represents the brightness profile obtained after processing with the P32 algorithm, while the dotted line shows the profile derived from identically processed data, but without correction for the transientresponse of the detector.
Open with DEXTER

$\begin{figure} \par\includegraphics[scale=0.65,clip]{h4471f16.eps} \end{figure}$	Figure 16: A $27\times 27$ arcmin field containing the galaxy M 101, as mapped in the C100 filter, after processing with the P32 algorithm. The map sampling is $15\times 23$ arcsec.
Open with DEXTER