We examine the effect that the shape of the source brightness profile has on the magnitude fluctuations of images in quasar lens systems due to microlensing .
We do this by convolving a variety of accretion disk models ( including Gaussian disks , uniform disks , “ cones , ” and a Shakura-Sunyaev thermal model ) with two magnification maps in the source plane , one with convergence \kappa = 0.4 and shear \gamma = 0.4 ( positive parity ) , and the other with \kappa = \gamma = 0.6 ( negative parity ) .
By looking at magnification histograms of the convolutions and using chi-squared tests to determine the number of observations that would be necessary to distinguish histograms associated with different disk models , we find that , for circular disk models , the microlensing fluctuations are relatively insensitive to all properties of the models except the half-light radius of the disk .
Shakura-Sunyaev models are sufficiently well constrained by observed quasar properties that we can estimate the half-light radius at optical wavelengths for a typical quasar .
If Shakura-Sunyaev models are appropriate , the half-light radii are very much smaller than the Einstein rings of intervening stars and the quasar can be reasonably taken to be a point source except in the immediate vicinity of caustic crossing events .