3 Optical observations

  
3.1 Imaging

Both target fields were observed with the Wide-Field Camera (WFC) on the 2.5 m INT telescope in dark time. Table 4 summarizes the main features of the images obtained. The WFC covers virtually all the field of view of EPIC, if centered optimally. In our case, all sources listed in Table 3 were imaged.

 

 

Table 4: Details of the optical wide-field imaging performed with the WFC on the INT.

Target	Mkn 205	G133-69 Pos_2
Observation date	30-04-2000	25/26-07-2000
Seeing (arcsec in i')	1.5	1.1
i' limiting magnitude	22.1	23.0
u exposure time (s)	600	-
g' exposure time (s)	600	600
r' exposure time (s)	600	600
i' exposure time (s)	1200	1200
Z exposure time (s)	1200	1200

The WFC images were reduced using standard techniques including de-bias, non-linearity correction, flat fielding and fringe correction (in i' and Z). Bias frames and twilight flats obtained during the same observing nights were used, but for the fringe correction contemporaneous archival i' and Z fringe frames were utilised. Information on the WFC pipeline procedures, which performs all these steps can be found under the Cambridge Astronomy Survey Unit (CASU) web pages.

The photometric calibration was performed by assuming average extinction constants and archival zero-point constants obtained routinely with the WFC, rather than measuring both from photometric standards.

Astrometric calibration was performed in two different ways. First a 12 parameter plate solution was applied to each one of the 4 CCDs of the WFC independently, by matching sources found in the image to USNO source positions. Residuals were typically found to be of the order of 0.3 arcsec for $\sim$100-200 matched sources. A second astrometric calibration (along the lines of the WFC survey) was performed by a simple 6 parameter plate solution over the whole 4-CCD image, but taking into account the previously calibrated telescope distortion cubic term. In this case the sources found were matched to APM source positions with similar residuals. We found the position of the candidate counterparts to X-ray sources to be consistent in both methods to within <0.5 arcsec.

3.2 Selection of candidate counterparts

In order to search for candidate counterparts of the X-ray sources, we have used the i'-band WFC image. Optical source lists for these images have been constructed by using the Sextractor algorithm (Bertin & Arnouts 1996). Sources were recorded if a minimum number of 10 connected pixels (of 0.33 arcsec) lay above 2 standard deviations of the background. Indeed some very faint sources escaped this detection algorithm, but the impact on the search for candidate counterparts was very limited.

Counterparts for the X-ray sources were searched for in the optical image lists. Candidate counterparts had to be either within the 5$\sigma$ statistical error or within 5 arcsec from the position of the X-ray source. This last criterion was used to accommodate any residual systematics in the astrometric calibration of the X-ray EPIC images.

The result of this search was very encouraging. Out of the 29 X-ray sources, 24 showed a single candidate counterpart, 4 showed two candidate counterparts and the remaining source showed no or very faint candidate counterpart in the i'-band image. The two basic conclusions to extract from this are that the selection of filter and depth of the images are appropriate to this sample and that the X-ray positional errors provided by XMM-Newton are good enough to single out a unique counterpart for the vast majority of the high-galactic latitude sources.

3.3 Spectroscopic observations

Candidate counterparts were spectroscopically observed with various spectrographs in order to identify the nature of the sources. Table 5 lists the relevant parameters of the spectroscopic instrumental setup used. At the source density we are dealing with ($\sim$ $100\deg^{-2}$) fibre spectrographs are likely to be the most efficient. However, we had a very limited amount of nights with this instrumentation and furthermore the fibres used were too large (2.7 arcsec diameter) to detect the faintest objects. Therefore a significant fraction of the sources were identified with long-slit spectrographs.

 

 

Table 5: List of spectroscopic setups relevant to this sample.

Telescope	Instrument	Spectral	Slit width	Spectral	Comments
		range (Å)	(arcsec)	resolutiona (Å)
WHT	AUTOFIB2/WYFFOS	3900-7100	2.7b	6-7	Fibre
WHT	ISIS	3500-8500	1.2-2.0	3.0-3.3	Long slit
TNG	DOLORES	3500-8000	1.0-1.5	14-15	Long slit

^a Measured from unsaturated arc lines.
^b Width of individual fibres.

As pointed out before, the vast majority of X-ray sources have a unique candidate counterpart. For the cases where there were two candidate counterparts we observed the brightest one (which happened also to be the closest one to the X-ray source) which invariably turned out to be a plausible identification (i.e. some sort of AGN).

The spectra were reduced using standard IRAF techniques. These included de-biasing, flat fielding, illumination correction (whenever a twilight flat was available), cosmic ray rejection, spectral extraction and background subtraction, arc lamp wavelength calibration and flux calibration by using spectrophotometric standards. The flux calibration does not have to be understood in absolute terms, as significant fractions of the light escaped the corresponding apertures and no attempt has been made to correct for this. Nevertheless the overall spectral shape, if not the normalisation, should be approximately correct. That is especially true for the long-slit spectra which in the majority of the cases were obtained with the slit oriented in parallactic angle.

The case of fibre spectroscopy deserves further comment. The relative fibre throughput was obtained by observing in offset sky positions within the same fibre configuration under use. In the proper on-target observations, the sky was subtracted by combining all the sky fibres that we placed in regions free from bright sources. We applied the same procedure to the offset sky observations, which allowed us to tweak the relative sensitivity of the fibres by making sure that no significant residuals were left in these spectra after the sky was subtracted. In a couple of iterations we found a satisfactory solution for the relative fibre sensitivity and then we used this to sky-subtract the target apertures. It must be stressed that sky lines are very difficult to subtract at this spectral resolution and therefore residuals will unavoidably be present in the fibre spectra.

We must emphasize that fibre spectroscopy, even with the large fibre aperture of 2.7 arcsec, was very efficient in identifying the optically brightest conterparts in the Mkn 205 field. We allocated a total of 26 fibres on optical sources in a single setup, which included several corresponding to X-ray sources below the flux cutoff. We succeeded in identifying 8 X-ray sources above the flux cutoff and a futher 3 below it in a total time of 2.5 hours including all overheads. This is at least twice as efficient as long-slit spectroscopy on the same telescope.

This high efficiency in the fibre spectroscopy relied, however, on two basic facts: the night was dark (so the sky background and noise were as low as possible) and the number of X-ray sources with a single candidate counterpart was very large. It must be pointed out that fibre spectrographs do not usually allow to place fibres within several arcsec at best, and therefore it is not possible to observe different candidate counterparts of the same X-ray source within the same fibre setup.