Microlensing

Context. Observations of the effect of microlensing in gravitationally lensed quasars could potentially be used to study the structure of the source on very small distance scales – down to the size of the supermassive black hole, that powers the quasar activity. Aims. We search for the microlensing effect in the γ-ray band using the signal from a gravitationally lensed blazar, B0218+357. Methods. We develop a method to deconvolve the contributions from the two images of the source, which allows us to reconstruct the flaring light curve in the γ-ray band. We use this method to study the evolution of the magnification factor ratio between the two images throughout the flaring episodes. We interpret the time variability of the ratio as a signature of the microlensing effect and derive constraints on the physical parameters of the γ-ray source by comparing the observed variability properties of the magnification factor ratio with those derived from numerical simulations of the microlensing caustics networks. Results. We find that the magnification factor ratio experienced a change characteristic for a caustic-crossing microlensing event that occurred during a 100 d flaring period in 2012. It changed again between 2012 and a recent flaring episode in 2014. We use the measurement of the maximal magnification and duration of the caustic-crossing event to derive an estimate of the projected size of the γ-ray emission region in B0218+357, Rγ ~ 1014−1015 cm. This estimate is compatible with a complementary estimate found from the minimal variability time scale. The microlensing/minimal variability time scale measurements of the source size suggest that the γ-ray emission is produced at the base of the blazar jet, in the direct vicinity of the central supermassive black hole