The size of a quasar's mid-IR emission region inferred from microlensed images of Q2237+0305

Abstract

We use published mid-IR and V-band flux ratios for images A and B of Q2237+0305 to demonstrate that the size of the mid-IR emission region has a scale comparable to or larger than the microlens Einstein Radius (ER, ~10^17 cm for solar mass stars). Q2237+0305 has been monitored extensively in the R and V-bands for ~15 years. The variability record shows significant microlensing variability of the optical emission region, and has been used by several studies to demonstrate that the optical emission region is much smaller than the ER for solar-mass objects. For the majority of the monitoring history, the optical flux ratios have differed significantly from those predicted by macro-models. In contrast, recent observations in mid-IR show flux ratios similar to those measured in the radio, and to predictions of some lens models, implying that the mid-IR flux is emitted from a region that is at least 2 orders of magnitude larger than the optical emission region. We have calculated the likeli-hood of the observed mid-IR flux ratio as a function of mid-IR source size given the observed V-band flux ratio. The expected flux ratio for a source having dimensions of ~1 ER is a sensitive function of the macro model adopted. However we find that the probability of source size given the observed flux ratios is primarily sensitive to the ratio of the macro-model magnifications. The majority of published macro models for Q2237+0305 yield a flux ratio for images B and A of 0.8 - 1.1. By combining probabilities from the ratios A/B and C/D we infer that the diameter of a circular IR emission region is >1ER with >95% confidence. For microlensing by low-mass stars, this source size limit rules out non-thermal processes such as synchrotron as mechanisms for mid-IR emission.