next previous
Up: The ATESP radio survey


Subsections

  
3 The source catalogue

The procedures described in the previous section yielded a preliminary list of sources (or source components) for further investigation. In order to minimize the incompleteness effects present at fluxes approaching the source extraction threshold (see Fig. 1) we decided to insert in the final catalogue only the sources with $S_{\rm peak} \geq 6\sigma$, where $\sigma$ is the mosaic rms flux density. This threshold has been chosen after inspection of the local noise distribution. The local noise ( $\sigma_{\rm local}$) has been defined as the average noise value in a box of about $8\hbox{$^\prime$ }\times 8\hbox{$^\prime$ }$ size around a source. Usually the local noise does not show significant systematic departures from the mosaic average rms value: the $\sigma _{\rm local}/\sigma _{\rm fit}$ distribution can be described fairly well by a Gaussian with FWHM = 0.14 and peak value equal to 1.01 (see Fig. 2).

This can be seen also in Fig. 3, where we show, for each source, the signal-to-noise ratio defined using both $\sigma_{\rm local}$ and $\sigma_{\rm fit}$. The two signal-to-noise ratios mostly agree with each other, although a number of significant departures are evident for the faintest sources. This is due to the presence of some residual areas where the noise is not random due to systematic effects (noise peaks and stripes). This is caused by the limited dynamical range in presence of very strong sources (stronger than 50-100 mJy, see also Paper I). It is worth noting that also the systematic departures from the expected behavior at the brightest end of the plot ( $S_{\rm peak} / \sigma_{\rm local} <
S_{\rm peak} / \sigma_{\rm fit}$) are a consequence of the same problem.

In mosaic regions where local noise is significantly larger, we applied a $6\sigma_{\rm local}$ cut-off if $\sigma_{\rm local}/\sigma_{\rm fit}
\geq 1.2$ (assuming a normal distribution for the noise, the probability to get a local to average noise value $\geq 1.2$ is $\leq 0.1\%$). This resulted in the rejection of 32 sources (see the region in Fig. 3 defined by $S_{\rm peak} \geq 6\sigma _{\rm fit}$, $S_{\rm peak} < 6\sigma_{\rm local}$and $\sigma_{\rm local}/\sigma_{\rm fit}
\geq 1.2$).

The criteria discussed above proved to be very effective in selecting out noise artifacts from the catalogue. Nevertheless a few (6) sources, which satisfy both the average and the local noise constraints are, however, evident noise artifacts at visual inspection. Such objects have been rejected from the final catalogue.

The adopted criteria for the final catalogue definition also allowed us to significantly reduce the number of poor Gaussian fits (see Sect. 2.3) since a large fraction of them ($\sim 65\%$) have fainter $S_{\rm peak}$: we are left with 50 poor Gaussian fits (flagged "S*'', see Sect. 3.4) in the final catalogue.

A total number of 3172 source components entered the final catalogue. Some of them have to be considered different components of a unique source and, as discussed later in Sect. 3.2, the number of distinct sources in the ATESP catalogue is 2960.

  
3.1 Deconvolution


  \begin{figure}
\par\resizebox{8.8cm}{!}{\includegraphics{H2076f4.ps}}
\end{figure} Figure 4: Ratio of the integrated flux $S_{\rm total}$ to the peak one $S_{\rm peak}$ as a function of the source signal-to-noise. Dotted lines indicate the $6\sigma _{\rm fit}$ cut-off adopted for the catalogue definition (vertical line) and the $S_{\rm total}=S_{\rm peak}$ locus (horizontal line) respectively. Also shown are the lower and upper envelopes (dashed lines) of the flux ratio distribution containing $\sim 90\%$ of the unresolved sources (dots). Filled circles indicate extended sources

The ratio of the integrated flux to the peak flux is a direct measure of the extension of a radio source:

\begin{displaymath}S_{\rm total}/S_{\rm peak}=\theta_{\rm min} \; \theta_{\rm maj} /b_{\rm min} \; b_{\rm maj}
\end{displaymath} (1)

where $\theta_{\rm min}$ and $\theta_{\rm maj}$ are the source FWHM axes and $b_{\rm min}$ and $b_{\rm maj}$ are the synthesized beam FWHM axes. The flux ratio can therefore be used to discriminate between extended (larger than the beam) and point-like sources.

In Fig. 4 we have plotted the flux ratio as a function of the signal-to-noise for all the sources (or source components) in the ATESP catalogue. The flux density ratio has a skew distribution, with a tail towards high flux ratios due to extended sources. Values for $S_{\rm total}/S_{\rm peak} < 1$ are due to the influence of the image noise on the measure of source sizes (see Sect. 4). To establish a criterion for extension, such errors have to be taken into account. We have determined the lower envelope of the flux ratio distribution (the curve containing 90% of the $S_{\rm total}<S_{\rm peak}$ sources) and we have mirrored it on the $S_{\rm total}>S_{\rm peak}$ side (upper envelope in Fig. 4). We have considered as unresolved all sources laying below the upper envelope. The upper envelope can be characterized by the equation:

\begin{displaymath}S_{\rm total}/S_{\rm peak} = 1.05 +
\left[ \frac{ 10 }{ (S_{\rm peak}/\sigma_{\rm fit})^{1.5}}\right] \; .
\end{displaymath} (2)

From this analysis we found that 1864 of the 3172 sources (or source components) in the catalogue (i.e. $\sim 60\%$) have to be considered unresolved.

It is worth noting that the envelope does not converge to 1 going to large signal-to-noise values. This is due to the radial smearing effect. It systematically reduces the source peak fluxes, yielding larger $S_{\rm total}/S_{\rm peak}$ ratios (see discussion in Sect. 4.2). From Fig. 4 we can quantify the smearing effect in $\sim 5\%$on average.

Deconvolved angular sizes are given in the catalogue only for sources above the upper curve (filled circles in Fig. 4). For unresolved sources (dots in Fig. 4) deconvolved angular sizes are set to zero. Note that no bandwidth correction to deconvolved sizes has been applied. Correcting for such effect would be somewhat complicated by the fact that each source in the radio mosaics is a sum of contributions from several single pointings.

  
3.2 Multiple sources


  \begin{figure}
\par\resizebox{8.8cm}{!}{\includegraphics{H2076f5.ps}}
\end{figure} Figure 5: Nearest neighbor pair density distribution as a function of distance (arcsec) for the ATESP radio sources (histogram) and the expected one assuming a random distribution of the sources in the sky (solid line). The expected distribution has been scaled so as to have the same area below the curve and the observed histogram. The excess at small distances is due to physical associations, and is compensated by a deficiency at intermediate distances (at $d\simeq 80-300\hbox {$^{\prime \prime }$ }$). Edge effects explain the discrepancy between the expected distribution and the observed one at very large distances ( $d>400\hbox {$^{\prime \prime }$ }$). The vertical dashed line indicates the $d=45\hbox {$^{\prime \prime }$ }$ cut applied to discriminate between real and unreal physical associations


  \begin{figure}
\par\resizebox{8.8cm}{!}{\includegraphics{H2076f6.ps}}
\end{figure} Figure 6: Pair components' flux ratios as a function of the distance between the two components for all the $d\leq 150\hbox {$^{\prime \prime }$ }$ pairs in the catalogue. Filled circles indicate solely the pairs that have been considered as part of a unique multiple source in the final form of the catalogue (see text for the criteria adopted for multiple sources' definition)

In Fig. 5 the (nearest neighbor) pair density distribution is shown as a function of distance (histogram). Also indicated is the expected distribution if all the sample sources (components) were randomly distributed in the sky. The expected distribution has been scaled so as to have the same area below the curve and the observed histogram. The excess at small distances is clearly due to physical associations and, because of the normalization chosen, is compensated by a deficiency at larger distances (between $80\hbox{$^{\prime\prime}$ }$ and $300\hbox{$^{\prime\prime}$ }$).

All the components closer than $45\hbox{$^{\prime\prime}$ }$ (i.e. about three times the beam size) have been considered as possibly belonging to a unique double source. Triple sources are defined whenever one additional component is closer than $45\hbox{$^{\prime\prime}$ }$ to (at least) one of the pair components. For multiple sources the same criterion is applied iteratively. Applying this distance constraint we expect that $\sim 20\%$ of the pairs are random superpositions.

The flux ratio distribution between the pair components has a large spread at all distances (see Fig. 6). To reduce the contamination we have discarded all the pairs with flux ratio larger than a factor 10. For triples and multiple sources the probable core is not considered when computing the flux ratios.

A few departures from the adopted criteria are present in the catalogue. For example the triple source ATESP J005620-394145 and the double source ATESP J011029-393253 have $d\leq 45\hbox{$^{\prime\prime}$ }$ but do not satisfy the flux ratio constraint. All exceptions are based on source geometry considerations and/or the analysis of the source field.

In order to increase the multiple sources' sub-sample completeness, we added 31 sources with distances $45\hbox{$^{\prime\prime}$ }< d < 150\hbox{$^{\prime\prime}$ }$, which show clear signs of physical associations between their components (see Fig. 7 for some examples). No flux ratio constraints have been applied to such sources. In Fig. 6 are shown the flux ratios for all the pairs in the final sample of multiple sources (filled circles).

As a final result we have 189 multiple sources: 168 doubles, 19 triples and 2 sources with four components. As a consequence, the initial list of 3172 radio components results in a catalogue of 2960 distinct radio sources.

  
3.3 Non-Gaussian sources

In the final catalogue we have 23 non-Gaussian sources whose parameters have been defined as discussed in Sect. 2.3. In particular we notice that positions refer to peak positions, which, for non-Gaussian sources does not necessarily correspond to the position of the core. We also notice that we can have non-Gaussian components in multiple sources. Some examples of single and multiple non Gaussian sources are shown in Fig. 7.

  \begin{figure}
\par {\includegraphics[width=4.6cm]{H2076f7a.ps}\hspace*{1mm}
\in...
...7h.ps}\hspace*{1mm}
\includegraphics[width=4.6cm]{H2076f7i.ps} }
%
\end{figure} Figure 7: Examples of very large ( $d>45\hbox {$^{\prime \prime }$ }$, see text) multi-component ATESP sources and of non-Gaussian ATESP sources. For direct comparison $4\hbox {$^\prime $ }\times 4\hbox {$^\prime $ }$ contour images are presented for all sources. For each source, the contour levels are at 3, 6, 10, 20, 30, 50, 100, 300 times the average rms flux density of the mosaic where the source has been detected. 1$^{\rm st}$ and 2$^{\rm nd}$ row: double sources. 3$^{\rm rd}$ row: triple sources


 \begin{figure}\addtocounter{figure}{-1}
\par {\includegraphics[width=4.6cm]{H207...
...6f7s.ps}\hspace*{1mm}
\includegraphics[width=4.6cm]{H2076f7t.ps} }\end{figure} Figure 7: continued. Examples of very large ( $d>45\hbox {$^{\prime \prime }$ }$, see text) multi-component ATESP sources and of non-Gaussian ATESP sources. For direct comparison $4\hbox {$^\prime $ }\times 4\hbox {$^\prime $ }$ contour images are presented for all sources. For each source, the contour levels are at 3, 6, 10, 20, 30, 50, 100, 300 times the average rms flux density of the mosaic where the source has been detected. 1$^{\rm st}$ row: the two 4-component sources in the ATESP catalogue. 2$^{\rm nd}$ row: single-component non Gaussian sources. 3$^{\rm rd}$ row: double sources with one (or two) non-Gaussian components

  
3.4 The catalogue format

The electronic version of the full radio catalog is available through the ATESP page at http://www.ira.bo.cnr.it. Its first page is shown as an example in Table 1. The source catalogue is sorted on right ascension. The format is the following:

Column (1) - Source IAU name. Different components of multiple sources are labeled "A'', "B'', etc.
Columns (2) and (3) - Source position: Right Ascension and Declination (J2000).
Columns (4) and (5) - Source peak ( $S_{\rm peak}$) and integrated ( $S_{\rm total}$) flux densities in mJy (Baars et al. 1977 scale). The flux densities are not corrected for the systematic effects discussed in Sect. 4.2.

Columns (6) and (7) - Intrinsic (deconvolved from the beam) source angular size. Full width half maximum of the major ( $\Theta_{\rm maj}$) and minor ( $\Theta_{\rm min}$) axes in arcsec. Zero values refer to unresolved sources (see Sect. 3.1 for more details).
Column (8) - Source position angle (PA, measured N through E) for the major axis in degrees.
Column (9) - Flag indicating the fitting procedure and parameterization adopted for the source or source component (see Sects. 2.3 and 3.2). "S'' refers to Gaussian fits. "S*'' refers to poor Gaussian fits. "E'' refers to non-Gaussian sources. "M'' refers to multiple sources (see below).

 

 
Table 1: The radio catalogue: first page
IAU Name RA DEC $S_{\rm peak}$ $S_{\rm total}$ $\Theta_{\rm maj}$ $\Theta_{\rm min}$ PA  
  (J2000) mJy arcsec degr.  
ATESP J223235-402642 22:32:35.62 -40:26:42.1 1.26 6.07 27.13 13.55 28.9 S
ATESP J223237-393113 22:32:37.71 -39:31:13.2 0.69 0.71 0.00 0.00 0.0 S
ATESP J223238-394102 22:32:38.54 -39:41:02.9 0.48 0.58 0.00 0.00 0.0 S
ATESP J223242-393054 22:32:42.10 -39:30:54.3 2.33 2.46 0.00 0.00 0.0 S
ATESP J223248-401345 22:32:48.24 -40:13:45.9 0.95 0.83 0.00 0.00 0.0 S
ATESP J223250-394059 22:32:50.79 -39:40:59.7 0.57 0.93 10.47 0.00 -88.4 S
ATESP J223252-401925 22:32:52.45 -40:19:25.1 0.66 0.46 0.00 0.00 0.0 S
ATESP J223254-393652 22:32:54.54 -39:36:52.0 1.03 1.70 9.89 5.39 48.3 S
ATESP J223255-395717 22:32:55.51 -39:57:17.3 0.51 0.45 0.00 0.00 0.0 S
ATESP J223256-402010 22:32:56.10 -40:20:10.3 1.13 1.08 0.00 0.00 0.0 S
ATESP J223301-393017 22:33:01.55 -39:30:17.1 7.83 8.67 3.39 2.66 -57.0 S
ATESP J223302-402817 22:33:02.31 -40:28:17.5 0.66 0.71 0.00 0.00 0.0 S
ATESP J223303-401629 22:33:03.07 -40:16:29.2 0.99 1.27 5.30 4.74 62.9 S
ATESP J223304-395639 22:33:04.92 -39:56:39.0 1.85 2.94 23.51 - - M
ATESP J223304-395639A 22:33:04.43 -39:56:41.4 1.85 2.18 6.66 1.43 -24.2 S
ATESP J223304-395639B 22:33:06.31 -39:56:32.2 0.80 0.76 0.00 0.00 0.0 S
ATESP J223313-400216 22:33:13.42 -40:02:16.1 7.60 9.18 4.94 3.65 -45.7 S
ATESP J223314-394942 22:33:14.42 -39:49:42.8 1.28 1.33 0.00 0.00 0.0 S
ATESP J223316-393124 22:33:16.90 -39:31:24.4 0.52 0.68 0.00 0.00 0.0 S
ATESP J223317-393235 22:33:17.19 -39:32:35.0 2.40 3.37 9.29 4.22 9.4 S
ATESP J223320-394713 22:33:20.14 -39:47:13.8 0.96 1.08 0.00 0.00 0.0 S
ATESP J223322-401710 22:33:22.93 -40:17:10.7 4.92 6.26 7.56 3.34 14.7 S
ATESP J223327-395836 22:33:27.45 -39:58:36.9 3.24 5.16 13.60 3.39 -5.9 S
ATESP J223327-394541 22:33:27.73 -39:45:41.7 2.74 5.70 21.21 - - M
ATESP J223327-394541A 22:33:27.08 -39:45:40.2 2.74 3.65 5.98 4.83 60.7 S
ATESP J223327-394541B 22:33:28.89 -39:45:44.4 1.29 2.05 10.04 0.00 -73.8 S
ATESP J223329-402019 22:33:29.10 -40:20:19.6 1.41 1.82 6.85 3.75 -32.1 S
ATESP J223330-395233 22:33:30.97 -39:52:33.4 0.50 0.53 0.00 0.00 0.0 S
ATESP J223335-401337 22:33:35.21 -40:13:37.7 0.63 0.71 0.00 0.00 0.0 S
ATESP J223337-394253 22:33:37.01 -39:42:53.8 2.58 2.79 0.00 0.00 0.0 S
ATESP J223338-392919 22:33:38.49 -39:29:19.6 5.26 5.62 0.00 0.00 0.0 S
ATESP J223339-393131 22:33:39.86 -39:31:31.0 1.37 1.32 0.00 0.00 0.0 S
ATESP J223343-393811 22:33:43.21 -39:38:11.2 1.00 1.05 0.00 0.00 0.0 S
ATESP J223343-402307 22:33:43.93 -40:23:07.4 0.90 1.10 0.00 0.00 0.0 S
ATESP J223345-402815 22:33:45.23 -40:28:15.6 0.56 0.65 0.00 0.00 0.0 S
ATESP J223346-393322 22:33:46.35 -39:33:22.9 1.26 1.23 0.00 0.00 0.0 S
ATESP J223351-394040 22:33:51.19 -39:40:40.9 1.12 1.17 0.00 0.00 0.0 S
ATESP J223356-401949 22:33:56.57 -40:19:49.7 1.71 2.28 36.37 - - M
ATESP J223356-401949A 22:33:55.33 -40:20:11.8 0.52 0.63 0.00 0.00 0.0 S
ATESP J223356-401949B 22:33:57.05 -40:19:41.2 1.71 1.65 0.00 0.00 0.0 S
ATESP J223358-400642 22:33:58.77 -40:06:42.4 1.29 1.20 0.00 0.00 0.0 S
ATESP J223401-402310 22:34:01.25 -40:23:10.8 0.58 1.18 10.98 9.14 32.8 S
ATESP J223401-393448 22:34:01.87 -39:34:48.9 1.69 2.09 5.28 3.10 -72.8 S
ATESP J223402-402357 22:34:02.36 -40:23:57.7 0.59 0.57 0.00 0.00 0.0 S
ATESP J223402-400017 22:34:02.64 -40:00:17.3 1.70 1.99 5.37 2.86 -28.2 S
ATESP J223404-395831 22:34:04.00 -39:58:31.5 6.16 7.45 5.53 3.43 29.7 S
ATESP J223404-393358 22:34:04.08 -39:33:58.9 0.85 0.90 0.00 0.00 0.0 S
ATESP J223407-393721 22:34:07.34 -39:37:21.9 3.14 6.41 25.71 - - M
ATESP J223407-393721A 22:34:06.35 -39:37:27.4 2.03 3.23 8.90 6.08 38.9 S
ATESP J223407-393721B 22:34:08.36 -39:37:16.3 3.14 3.19 0.00 0.00 0.0 S
ATESP J223409-394258 22:34:09.58 -39:42:58.2 0.71 1.11 10.16 5.23 -15.4 S
ATESP J223410-394427 22:34:10.82 -39:44:27.6 1.54 1.59 0.00 0.00 0.0 S
ATESP J223412-400254 22:34:12.80 -40:02:54.6 0.66 0.69 0.00 0.00 0.0 S
ATESP J223413-393242 22:34:13.28 -39:32:42.4 1.58 1.88 6.21 0.00 59.9 S
ATESP J223413-393650 22:34:13.74 -39:36:50.2 0.48 1.71 25.27 8.28 38.4 S
ATESP J223413-395651 22:34:13.80 -39:56:51.7 0.52 0.56 0.00 0.00 0.0 S
ATESP J223420-393150 22:34:20.47 -39:31:50.5 0.55 0.73 0.00 0.00 0.0 S


The parameters listed for non-Gaussian sources are defined as discussed in Sect. 2.3.

For multiple sources we list all the components (labeled "A'', "B'', etc.) preceded by a line (flagged "M'') giving the position of the radio centroid, total flux density and overall angular size of the source. Source positions have been defined as the flux-weighted average position of all the components (source centroid). For sources with more than two components the centroid position has been replaced with the core position whenever the core is clearly recognizable.

Integrated total source flux densities are computed by summing all the component integrated fluxes.

The total source angular size is defined as las (see Sect. 2.3) and it is computed as the maximum distance between the source components.


next previous
Up: The ATESP radio survey

Copyright The European Southern Observatory (ESO)