Aims. We considered the possibility of measuring time delays between components of the multiplied quasar QSO2237+0305 and between V and R band variations. Methods. The analysis is based on the flux variations of four components observed by the OGLE collaboration and Maidanak group during the last quarter of 2003. The observed gradients of the brightness variations in this period for OGLE data in V band are 3.4 mmag/day, 2.2 mmag/day, 2.4 mmag/day, and 1.1 mmag/day for the components A, B, C, and D, respectively. The variations are probably intrinsic source variations. The basic method used for time-delay measurements is the cross-correlation technique. To estimate the uncertainty of the time-delay measurements, Monte Carlo simulations were carried out. Results. The calculations showed the impossibility of unambiguously measuring the differential time delays between the components. The observations at shorter wavelengths seem to be the only way to achieve robust time delay measurements in the system. The wavelength-dependent time delays can be used to provide the evidence of an accretion-disk structure of the central optical variable source. The observed V and R band variations of component C show good correlation with the correlation coefficient of 0.83. However, the obtained time delay, about 16.2 days, and its accuracy are far from reliable.

We have observed the gravitational lens system Q2237+0305 from the Maidanak Observatory over the period from 2002 August to 2003 November. Here we report the results of our observations. We implemented a two‐stage technique that has been developed specifically for the purpose of gravitational lens image reconstruction. The technique is based on the Tikhonov regularization approach and allows one to obtain astrometric and photometric characteristics of the gravitational lens system. Light curves with 78 data points for the four quasar components are obtained. Slow brightness variations over the observational period are found in all components. Images A, C and D have a tendency to decrease in brightness. Image B does not vary more than 0.05 mag. The observations did not reveal evidence for large variations in brightness of the components due to microlensing effects. To provide an overall picture of the photometry behaviour, our data are combined with the Maidanak observations published for 1995–2000.

ABSTRACT. We show that, contrary to simple predictions, most AGNs show at best only a small increase of lags in the J, H, K, and L bands with increasing wavelength. We suggest that a possible cause of this near simultaneity of the variability from the near-IR to the mid-IR is that the hot dust is in a hollow bi-conical outflow of which we only see the near side. Although most AGNs show near simultaneity of IR variability, there was at least one epoch when NGC 4151 showed the sharply increasing IR lag with the increase of the wavelength. This behaviour might also be present in GQ Comae. We discuss these results briefly. The relative wavelength independence of IR lags simplifies the use of IR lags for estimating cosmological parameters.

The microlensing high‐amplification events in the light curves of the gravitationally lensed quasar Q2237+0305 observed by the OGLE group and GLITP collaboration are analysed. The significant brightness amplification in the A and C components in 1999 observational season are considered under the assumption of the fold caustic crossing. Under this assumption we applied the model‐independent method based on regularization technique for one‐dimensional profile restoration of the quasar accretion disc brightness distribution. The recovered brightness distribution of the source seems to obey the standard model for the accretion disc. The estimated effective radius of the quasar emitting region is in agreement with the previous studies.

No abstract