Astron. Astrophys. 347, 434-441 (1999)

5. Conclusion

The analysis of new data and the re-discussion of older ones from HST presented in this study lead to the following main conclusions:

• 1) Good resolution imaging at visible and IR wavelengths confirms the existence and precises the amplitude of differential extinction between the images of the quasar nucleus.

• 2) A reanalysis of HST data shows that the arc should be carefully taken into account for a correct photometry of these nucleus images. After doing so, the observed properties of the nucleus images become compatible with a classical extinction by dust.

• 3) It is not yet possible to make a clear departure between the contributions of the source and of the lensing galaxy to the global extinction. Fitting the observed spectrum with dust at both redshifts would certainly work well, but the quality of present data is not sufficient for putting strong constraints on the fit parameters.

• 4) The source was not significantly variable during the period 1994-1997.

Concerning the last point, one may note that, compared to earlier observations in 1991-1992 (Angonin-Willaime et al., 1994), a significant ( 0.5 magnitude) dimming is observed in R for the close pair (A1+A2) while B and C remain constant. In I, there is no mesurable effect, save for a small global shift of the zero point that is imputable to passband differences (see the discussion in Sect. 2.2). This could be the signature of an intrinsic change in the quasar combined with the different time delays of the images. Another possible explanation is that we were witnessing in 1991-92 the termination of a microlensing event that ended before 1994. However, the difference in behaviour between I and R is not easily explained with such mechanisms.

The simplest explanation for this chromatic effect is that we are in presence of time-dependent absorption of the image(s) by the lens. The time scale seems a priori inadequate, but the lensing would help to shrink it. A1-A2 is a very amplified pair of images (A 20 to 30 in flux) near a caustic. This flux amplification is only a geometrical effect (magnification matrix of the lens), which acts on the vectors of relative proper motion as well as on position vectors. Statistically, the modulus of a randomly oriented proper motion vector should be increased by , with possibly higher amplifications for favourable orientations. In the specific case of the pair A1-A2 in MG J0414+0534, the beam of light coming from the source may thus scan the lens galaxy with a relative velocity of several thousands, or even tens of thousands of km s^-1. The same line of reasoning used for discussing the time scale of microlensing events holds also for the time scale of extinction variations across small clouds. If the clouds have a fine patchy structure, variable extinction may be observed with a time scale compatible with the data.

© European Southern Observatory (ESO) 1999

Online publication: June 30, 1999
helpdesk.link@springer.de