Some of the distant quasars are occasionally situated behind a
foreground galaxy. In this case the gravitational field of the galaxy
bends the light from the quasar so that the entire galalaxy works as cosmic
scale lense, producing multiple images of the quasar. This is the case
for the quasar PKS 1830-211 in which the effect of gravitational lensing
splits the image of the quasar into two images separated by the angular
distance of one arcsecond. The angular resolutions of gamma-ray
telescopes, like ESA's INTEGRAL and NASA's Fermi telescopes is only
about a fraction of a degree, which is not sufficient to resolve the two
images of the source. Still, the gamma-ray telescopes manage to separate
the flux from the two images, based on the variability properties of the
source. Differences in the path of photons coming from the two images
lead to a delay of the signal coming from one of the images by
approximately twenty days. Observing this time delay effect, astronomers
have noticed a peculiarity of the source variability pattern. The
advanced and delayed signals did not exactly repeat each other.
A consistent explanation of this peculiar behaviour is provided by the
effect of gravitational "microlensing". This effect is produced by stars
in the lensing galaxy. Each star passing through the line of sight
toward the quasar works as a gravitational lens. This also leads to the
appearance of the multiple images of the source, but the images are
separated by only about a micro-arcsecond. This angular scale could not
be resolved even by the best telescopes working in the visible domain, to
say nothing about the gamma-ray telescopes. However, the gamma-ray
telescopes are able to see the effect of these images on the variability
of the light from the source.
The variability arises because of the motion of the stars across the
line of sight. This motion effect is explained in the set of figures
A-D. Left top panel A shows a simulated pattern of magnification of the
source flux by many stars for different source positions. Red color
indicates the positions where the effect of the microlensing is strong, blue
is where the effect is moderate or absent. The "spider web" like
structure is a network of the microlensing "caustics". The flux of a
source situated exactly behind the caustic line is strongly magnified.
Relative motion of the stars (the caustics) and the source (the white
cross in the left top panel A) causes appearance / disappearance of
multiple images, as shown in the right top panel B. Appearance of the
new images and their stretching leads to strong magnification of the
flux at the moments when the source passes just behind the caustic. This
"caustic crossing" events are visible on the red dotted lightcurve in the
left bottom panel C. It is the microlensing effect which influences the
exceptionally bright flare of PKS 1830-211 observed by the Fermi/LAT
telescope (the data shown in the right bottom panel D). Using the
microlensing effect, the authors of the Paper
"Central engine of a gamma-ray blazar resolved through the magnifying glass of
gravitational microlensing"
have managed to
constrain the size of the source of gamma-rays detected by Fermi/LAT
from PKS 1830-211 and locate it very close to the black hole, at the
base of the AGN jet.
The effect of the microlensing is strong only if the source is very
small, much smaller than the typical caustic-to-caustic distance in the
left top panel A. The lightcurve of a large source (blue lighturve in
the panels C and D) does not reveal the caustic crossing episodes.
Non-observation of the caustic crossings in the INTEGRAL lightcurve
(shown in the right bottom panel D) shows that the hard X-ray / soft
gamma-ray source is large.
This large size resolves the puzzle of how high-energy gamma-rays can
actually escape from the compact source and be detected by Fermi/LAT, as
otherwise the high-energy gamma-rays would produce electron positron
pairs in interaction with the hard X-ray photons.
References:
"Central engine of a gamma-ray blazar resolved through the magnifying
glass of gravitational microlensing"
Andrii Neronov, Ievgen Vovk, Denys Malyshev
Nature Physics (August 2015, Volume 11 No 8 pp 664-667)
http://dx.doi.org/10.1038/nphys3376