Beffects

A visual and possibly way of testing how baryons affect the DM distribution is a direct comparison of the properties of arcs/multiple images in galaxies and clusters.

DM distributions in galaxies and clusters from num. simulations are very self-similar. However, if one compares the lensed images in galaxies with those in clusters, the properties are very very different and galaxies show much "smoother" behaviour. For example, why are galaxy arcs/rings not affected by (CDM) substructure in the same way as cluster arcs are affected by galaxy (their equivalent substructures)?

This either implies that galaxies don't have the CDM subtructures, their numbers are much less abundant as compared to clusters, or they are destroyed somehow in the proicess of galaxy formation (because they are less bound due to the absence of baryons in their DM potentitals).

Lensing, I think, can provide exactly the probe needed to test whetehr the amount of substructure in clusters and galaxies is what is expected from simulations and if not, lead to a modification of either the physics going in to the models and the physics itself (DM properties) **[LEON]**

Marusa: I agree this is potentially a very strong test of the substructure. However I think different apperances of arcs/multiple images in galaxies/clusters are by themselves not yet conclusive. One shouldn't forget that the relative sizes of sources as compared to the typical caustic sizes are very different in these two cases.

Therefore, do we know of good enough simulations that might be able to exactly tell us what we are suppose to see (in different scenarios?

Ole: I agree with you, Marusa, regarding the ratio of source sizes to lens mass being the main effect for the different behaviours. Simulations do predict this qualitatively. To make more quantitative predictions probably needs simulations that accurately take account of baryons and have high enough resolutions.