2 Observation and data reduction

The observation by the 60-cm space telescope ISO took place towards the center position of $l=-18.63\hbox{$^\circ$ }$ and $b=+0.35\hbox{$^\circ$ }$, i.e. ${\rm \alpha=16^{h}50^{m}25.4^{s}~ and ~\delta=-43\hbox{$^\circ$ }58\hbox{$^\prime$ }29\hbox{$^{\prime\prime}$ } }$. The rasters covered a rectangular $l \times b$ area of $0.35 \times 0.29$ deg². The details of the ISOGAL observation procedure with ISOCAM (Cesarsky et al. 1996) are described in Schuller et al. (2003). The observations of the field FC-01863+00035 were performed with 6 $\hbox{$^{\prime\prime}$ }$ pixels with the filters LW2 (5.0-8.5 $\mu$m, $\lambda_{\rm ref}=6.75~\mu$m) and LW3 (12-18 $\mu$m, $\lambda_{\rm ref}=14.3~\mu$m). The log of ISOGAL and DENIS observations is given in Table 1.

 

 

Table 1: Log of observations.

	Filter	Date	ISO ION	pfov
ISOGAL	LW2	1997.03.18	48801636	6 $^{\prime\prime}$
ISOGAL	LW3	1996.09.26	31500236	6 $^{\prime\prime}$
DENIS	I, J, $K_{\rm S}$	1996.03.25		3 $^{\prime\prime}$

The general processing of the ISOGAL data is described in detail in Schuller et al. (2003). The reduction of the data from the OLP7.0 (OffLine Processing) pipeline was performed by first using the CIA package (Ott et al. 1997). Dark currents are corrected and cosmic-rays are removed. Thereafter, a procedure is applied to simulating the time behavior of the pixels of the ISOCAM detectors, eliminating the related artifacts and improving the photometry. The flat-field and image distortion are corrected. Conversion to magnitudes from the calculated flux density at 7 $\mu$m and 15 $\mu$m is under the assumption that a Vega model, without circumstellar dust, corresponds to zero magnitude at respective wavelengths, i.e. ${[7]=12.38-2.5\log F_{\rm LW2}\rm (mJy)}$ and ${[15]=10.79-2.5\log F_{\rm LW3}\rm (mJy)}$. The current version of the images are shown in Fig. 1. On average, the rms dispersion of the ISOGAL photometry (from repeated observations - see for example Schuller et al. 2003) is estimated to be generally less than 0.2 mag with a small increase for the faintest sources that correspond to about the 50% completeness limit.

Figure 1: The ISOCAM images of the field FC-01863+00035 in LW2 (left) and LW3 (right) bands. Increasing l is to the right and increasing bis downwards. The sources with [7]<5.0 or [15]<5.0 are labeled by the sequence numbers in the ISOGAL PSC catalog. (See Table 4 for those detected by IRAS.)

In the field FC-01863+00035, the numbers of point sources extracted are 538 and 389 respectively in LW2 and LW3 bands within the limits of magnitude 9.38 in LW2 and 8.16 in LW3, which correspond to the flux limits of about 15 mJy and 11 mJy, respectively (generally such limits are chosen in the ISOGAL PSC such that they correspond to detection completeness $\sim$50%, Schuller et al. 2003). There are in total 648 ISO sources, out of which 279 objects were detected in both LW2 and LW3 bands, 259 objects detected in only LW2 band and 110 objects detected in only LW3 band. Among the 279 LW2-LW3 associated sources, 257 sources have good association quality flags 3 or 4 and 21 sources have doubtful associations with quality flag 2. We will discard the single association with quality flag 1 hereafter. With the association radius 5.4 $^{\prime\prime}$, the number of LW2-LW3 spurious associations should be $\sim$3. The ISO sources are distributed along brightness as shown in Fig. 2. As can be seen from Fig. 2 the detection is certainly not complete to sources fainter than magnitude 9.0 in LW2 or fainter than magnitude 7.5 in LW3. A general tendency is that more brighter sources are detected in both LW2 and LW3 bands while more weaker sources are detected in only one band.

Figure 2: The sources detected in the ISOGAL observation. The solid line histogram shows the distribution of all ISOGAL sources, the short dash line shows the distribution of the sources associated with the other ISOGAL band and the dash-dot line shows the distribution of the sources detected during the DENIS survey where only the sources with good association flags are counted. The left graph is the histogram of LW2 sources and the right one is of LW3 sources.

In near-infrared, this ISOGAL field was also observed in the 2MASS survey in the J, H and $K_{\rm S}$ bands, but the data are not yet available. Contrarily, it was early observed in a special observation of the DENIS survey in the I, J, $K_{\rm S}$ bands by a 1-m telescope at ESO, La Silla (Simon, in preparation). The DENIS survey has much higher sensitivity (by typically 3 mag in $K_{\rm S}$) as well as higher spatial resolution (by about 2- 3 times) than the ISOGAL survey. The number of sources detected by DENIS is much larger and limited by confusion in the J and $K_{\rm S}$ bands. There are 5345, 5817 and 5702 sources in the I, J and $K_{\rm S}$ bands, respectively. From the distribution of the brightness of these DENIS sources in this field as shown in Fig. 3, the detections in the near IR bands are reasonably complete to I=16 mag, J=14 mag and $K_{\rm S}=12$ mag.

In order to keep the population of spurious DENIS-ISOGAL cross-identifications below a few percents, the association was limited to $K_{\rm S}$-detected sources with $K_{\rm S}$ < 12.9 (shown by a long-dash line in the right panel of Fig. 3), corresponding to a density of $K_{\rm S}$ sources ${n=36~000/\rm deg^2}$. Such a density limit is systematically applied in the ISOGAL PSC for the DENIS association radius. It is chosen such as ${n\pi r_{\rm a}^2=0.1}$, where ${r_{\rm a}=3.5\hbox{$^{\prime\prime}$ }}$ is the main association radius. The numbers of sources associated with a DENIS source in the ISOGAL PSC are thus 479 objects in the LW2 band and 270 objects in the LW3 band. Among these objects, 16 LW2 sources and 18 LW3 sources (in total 24 sources only) are poorly associated with the DENIS objects, i.e. their association quality factors are either 1 or 2 in the ISOGAL PSC (see for details from Schuller et al. 2003) and they are dropped for later discussion. The numbers of associations with quality flag equal to 3 are 9 in the LW2 band and 10 in the LW3 band (15 sources in total); they mainly correspond to association radius between ${r_{\rm a}}$ and 2 ${r_{\rm a}}$. The associations with quality factors 4 or 5 (463 sources) are very probably real associations with a proportion of spurious associations less than $\sim$2%. The associations with quality flag 3 still have a good chance to be real, but with a larger proportion of spurious associations; in this field however no such objects have association radius larger than 5 $^{\prime\prime}$. Among the 538 LW2 objects, 463 (86% of all LW2 sources) are thus reasonably well associated with the DENIS $K_{\rm S}$sources, 351 are also detected in the J-band and 109 in all DENIS bands. Among the 389 LW3 objects, 252 (65% of all LW3 sources) are reasonably well associated with the DENIS $K_{\rm S}$ sources, 198 are also detected in the J-band and 54 in all DENIS bands. Three DENIS sources associated with ISOGAL have a bad quality in $K_{\rm S}$ because of saturation ( $K_{\rm S}<6.5$), and one J-associated source is saturated (J<8.0). One can check on the DENIS $K_{\rm S}$ and J images that five strong sources present at the position of strong ISOGAL sources are missing in the DENIS catalog because of saturation (Sect. 4.3). From Fig. 3 it can be seen that most of the DENIS objects with $K_{\rm S}< 9$ with 11 exceptions were detected by ISO. Among the ISOGAL sources that are not associated with the DENIS sources, 31 are detected in both LW2 and LW3 bands, 27 in only LW2 band and 87 in only LW3 band. There are 15 LW3 objects associated with $K_{\rm S}$ objects and with association quality flag $\geq$3, but are not associated with LW2 sources. They may be spurious cross-identification between DENIS and ISOGAL catalogues and are dropped in the following discussions.

The details of the results including the astrometric, photometric and association information of all the ISOGAL objects will be available on the web in the ISOGAL-DENIS PSC via http://vizier.u-strasbg.fr/viz-bin/VizieR and http://www-isogal.iap.fr (Schuller et al. 2003).

Figure 3: The DENIS sources in the ISOGAL field. All the sources detected by DENIS in the same field are distributed as indicated by the solid line. Those associated with the ISOGAL objects are represented by the short-dashed line and their number are also shown in the upper part of the graphs. From left to right are aligned bands I, J and $K_{\rm S}$. In the graph of the $K_{\rm S}$-band, a long-dashed vertical line marks the limit $K_{\rm S}=12.9$ mag used for searching DENIS association.