Astron. Astrophys. 328, 702-705 (1997)

2. Observations and data reduction

Five fields in low cirrus regions were mapped in raster mode (called `AOT P22') with stepsize using the ISOPHOT 25 filter and aperture (Lemke et al. 1996). Central positions of the fields and sizes of the maps are listed in Table 1. M01, M02, M03 refer to low background positions at ecliptic latitudes of = , , respectively. NGP stands for a field near the North Galactic Pole and ECL90,0 for a field in the ecliptic plane to be observed when the distance from the Sun is about .

Table 1. Coordinates and sizes of the mapped fields

Data reduction was performed using the ISOPHOT Interactive Analysis (PIA) version 6.0, including correction for non-linearities of the electronics, subtraction of dark current, and removal of cosmic ray hits (in some measurements the latter was performed by visual inspection of the integration ramps). The data were calibrated by using the default responsivity of 0.46 A/W of the P2 detector, and applying a preliminary correction of 0.55 for signal increase in the 180 aperture with respect to the aperture for which the calibration of responsivity had been performed on standard stars (see Lemke 1997). Comparing the average surface brightness of our maps, listed in Table 2, with COBE/DIRBE weekly maps (see Boggess et al. 1992 and Silverberg et al. 1993) we found deviations of less than 10% (no colour correction has been applied to either data set).

Table 2. Instrumental noise and fluctuation of surface brightness [in MJy/sr]

Because long term detector drifts in ISOPHOT were anticipated, we also performed two cross scans on each map following the raster measurement. These short duration scans are less affected by drift, providing a possibility to correct the raster measurement by comparing directly sky positions observed both in the cross scans and in the map. Also linear gradients were removed from the map. After this process, the map looks like shown in Figure 1. Since the initial part of the drift in the raster measurement could not be well corrected for by our procedure, we omitted the pixels around the edge of the field, where the first measured pixels are situated, in the following analysis.

The fluctuation of surface brightness in the maps was determined by two independent methods. First, we plotted the histogram of pixels and fitted the central part of the distribution by a Gaussian (see Figure 2). The standard deviation of this fit is denoted in Table 2 as `Histogram, 1 '. As a second method, we calculated the structure function of the maps,

Fig. 2. Map of x obtained at the North Galactic Pole. The data have been drift-corrected as described in the text. Note that the map shows full pixels for convenience only - the actual measurements were performed with a circular diaphragm of diameter.

where, for a given separation , the average is taken over all sky locations x of the map, and in addition was performed over all position angles of the separation vector. We then estimated the fluctuation as (see Figure 3). In general these two measures of noise agree. Table 2 lists the average surface brightness of the maps as well as the 1 sigma fluctuations in MJy/sr determined from the histogram and the structure function, respectively. We also included the instrumental measurement uncertainties, derived from the fluctuations between the individual brightness readings, as given by PIA. The last two columns give the fluctuation of the maps after subtracting this instrumental measurement uncertainty quadratically from the observed fluctuations. These columns constitute the primary result of our measurement.

Fig. 3. Histogramm of the brightnesses measured in the central 13x13 pixels of the map at the North Galactic Pole.

© European Southern Observatory (ESO) 1997

Online publication: March 26, 1998

helpdesk.link@springer.de