Is the Number of Giant Arcs in LCDM Consistent With Observations?

We use high-resolution N-body simulations to study the cross-sections and the predicted number of giant arcs in the LCDM model. We find that the prediction is sensitive to the source redshift, in qualitative agreement with Wambsganss et al. (2004). However, their assumption that the length-to-width ratio of arcs can be approximated by the magnification appears invalid. The violation of this assumption lowers the lensing probability by a factor of $\sim 10$. Thus, our predicted rate of giant arcs appears lower than the observed rate, although the precise discrepancy is unclear due to uncertainties both in theory and observations.