PMAS observations of the quadruple QSO HE 0435-1223

L. Wisotzki (1,2), T. Becker (1), L. Christensen (1), A. Helms (2), K. Jahnke (1), A. Kelz (1), M. M. Roth (1), S. F. Sanchez (1)

(1) Astrophysikalisches Institut Potsdam, (2) Universität Potsdam

Abstract

We present the first spatially resolved spectroscopic observations of the recently discovered quadruple QSO and gravitational lens HE 0435-1223. Using the Potsdam Multi-Aperture Spectrophotometer (PMAS), we show that all four QSO components have very similar but not identical spectra. We also introduce some novel techniques for the analysis of integral field spectroscopic observations of gravitationally lensed quasars in general.

Finally, we present a detection of the lensing galaxy, although this is close to the limits of the data. Comparing with a model galaxy spectrum, we obtain a redshift estimate of z_lens = 0.44 +- 0.02.

A full account of this work and the relevant references can be obtained in the recently completed paper by L. Wisotzki et al., Astronomy and Astrophysics 2003, in press.
A preprint is available as PDF file or gzipped PS file.

The Target

HE 0435-1223 was first discovered in the course of the Hamburg/ESO survey for bright QSOs (Wisotzki et al. 2000) and recently found to be a rather spectacular example of a quadruply imaged QSO (Wisotzki et al. 2002). The quasar is at a redshift of z = 1.689 and had a total magnitude g = 17.8 at the epoch of discovery, with evidence for significant variability. A composite colour image obtained with the 6.5m Magellan telescope is shown right. There can be no doubt about the nature of this object as being gravitationally lensed. However, so far there are no resolved spectra available of the individual components.

Observations

HE 0435-1223 was targeted with PMAS during the first regular (non-comissioning) observing run of the Potsdam Multi-Aperture Spectrophotometer (PMAS), mounted at the Calar-Alto 3.5 m telescope for several nights between 2--7 September, 2002. The image scale was 0.5 arcsec per spatial pixel, the total field of view was 8 arcsec x 8 arcsec. We used the V300 grating, giving a spectral resolution of 6A FWHM and a spectral range from 3950A to 7250A.

At the observing date in early September, the target became visible only very shortly before morning twilight, and only at airmasses of around 2. We revisited the object repeatedly and could record useful data at the end of three different nights, with a total exposure time of 6600 s. The external (DIMM) seeing was usually around or slightly below 1 arcsec. The data were reduced with the IDL-based package P3d.

Below we show three quasi-monochromatic images (bandwidth 3.3A), obtained at three different wavelengths. The orientation is standard (North is to the top, East is to the left). The four QSO components are clearly resolved. Notice the strong shift of image centroids as a function of wavelength, caused by differential atmospheric refraction.

lambda = 4278	lambda = 5601	lambda = 6924

Analysis

Considerations

The integral field approach allows the data to be accessed as either spectroscopic or as imaging data.

Long-slit approach: A direct extraction, spatial pixel by spatial pixel, will not allow spectrophotometry:
There are significant pixel shifts due to differential refraction (this is true even at low airmass).
There is significant mutual overlap of compenents at least in inner pixels.
Imaging approach: Fitting multiple PSFs to each monochromatic layer is also not without problems:
No PSF calibrator exists within the small field of view.
The PSF cannot be assumed to be constant with wavelength.

We developed a variant of the imaging approach which we describe in the following.

Extraction of spectra: Algorithm

Ansatz: Approximation of each QSO image as single elliptical Gaussian. --> 4 x 6 = 24 free parameters per slice. In order to reduce the number of free parameters, we derive constraints. Initial (a priori) constraints are:
- All point sources have same shape.
The relative image separations are known from the Magellan astrometry.

This is followed by a multi-step procedure:
1. Fit 4-component Gaussians with $2+3+4 = 9$ free parameters to each slice independently.
2. Replace x, y centroids by smooth polynomials (see figure).

3. Redo the fit, now with 3+4 = 7 parameters.
4. Replace position angle of PSF by mean value.
5. Redo fit, now with 2+4 = 6 parameters.
6. Replace fwhm(x) and fwhm(y) by smooth polynomials (see figure)
7. Final fit with only 4 amplitudes as parameters.

Results

The outcome of this algorithm is first displayed in the imaging domain below. We show the same three monochromatic slices as above, plus a simulated broad-band image (top row); the fit images for the same datasets (middle row); and the residuals data - fit (bottom row). Notice that the residuals are consistent with pure shot noise.

lambda = 4278 lambda = 5601 lambda = 6924 lambda = 5000-6000

Spectra

Now we show the extracted, calibrated and coadded spectra of the four QSO components. The green lines near zero represent the 1sigma error arrays. Notice that the S/N ratio of the spectra is quite high, up to 60 in the brightest component A. The adopted local continuum levels, used to derive emission line properties, are indicated by the red-dashed line segments.
The spectral slopes of components A, B, and D are indistinguishable, implying that extinction due to dust plays no major role in the lensing galaxy. On the other hand, the strengths of the emission lines relative to the continuum are quite different in the four components. We argue that most likely, microlensing is responsible for this phenomenon (see the full paper for details).

Line profiles

Below we show sets of pairwise comparison of normalised emission line profiles, obtained after subtracting a local continuum level. The upper row shows the C IV, the lower row the C III] lines. In each panel, the thick green lines represent component A, while the black thin lines denote components B, C, and D, respectively. Below each panel we show also the difference between the profiles.

The main conclusion here is that also the emission line profiles are identical within the error bars, exactly as expected from gravitational lensing. Furthermore, the identity of the line profiles makes it very unlikely that microlensing can be a major effect for the broad line region in this QSO.

Image Supersampling and Detection of The Lensing Galaxy

While the Magellan image shows the lensing galaxy quite prominently, a similarly clean detection could not be expected within the PMAS data, for three reasons:
1. The PMAS pixel grid is much coarser (0.5 arcsec for PMAS vs. 0.07 arcsec for Magellan+MagIC);
2. The effective seeing was clearly inferior (about 1.1 arcsec for PMAS vs. 0.6 arcsec for Magellan).
3. The nonavailability of a simultaneously obtained PSF reference made a clean PSF subtraction difficult.

We have nevertheless attempted to detect the lensing galaxy, both in `imaging' and in `spectroscopy' mode. In this we made use of some pieces of information available from the previous observations (Wisotzki et al. 2002).

In order to partly overcome the coarse pixel grid limitation, we have exploited the fact that due to atmospheric refraction, the QSO centroid gradually shifts as a function of wavelength. This can be seen as almost equivalent to extensive dithering with respect to a broad-band image. The image is shifted by several pixels over the entire wavelength range, a fortuitous byproduct of the high airmass during observation. We have written a small application to coadd monochromatic data planes into a broad-band image, including a 4 x 4 supersampling and applying the polynomial approximation to the centroid shifts described above.
The image to the right shows that the four point sources are indeed now much better defined than in the original coarse pixel data.

Guided by the redshift prior on the lensing galaxy, we selected only the red part of the spectrum, 5500A < lambda < 7200A. We then run also the residual image (data minus point sources) through the same procedure. The result is shown in the right-hand image.
We find that the strongest residual is always positive and occurs very close to the expected position of the lensing galaxy. Relative to the position of component A, the difference from the Magellan value is only 0.1 arcsec in right ascension and 0.2 arcsec in declination. Going through the same exercise with a lambda < 5500A image does not yield any discernible signal at that location. We therefore conclude that the detection is real and significant.

In order to obtain spectral information on the lensing galaxy, we extracted and coadded the spectra of a 2 x 2 pixels block (1 arcsec x 1 arcsec) at the expected location from each residual data cube. The result was an extremely noisy spectrum, which we then degraded in resolution by summing the data into bins of 300A width, with the intention to use the overall spectral energy distribution for a redshift estimate. This is documented the right-hand figure below.
We compared this with a model stellar population of age 10 Gyrs and cross-correlated the spectra. The left-hand panel of the figure below shows that chi2(z) has a strong minimum between z=0.43 and z=0.45. Since a lens redshift z>0.5 is not compatible with our prior assumptions, we settle at a redshift estimate of z_lens = 0.44 +- 0.02, for which we show our best fit model overplotted over the binned residual spectrum.