A Revised Model for the Formation of Disk Galaxies: Low Spin and Dark Halo Expansion

We use observed rotation velocity-luminosity (VL) and size-luminosity (RL) relations to single out a specific scenario for disk galaxy formation in the LambdaCDM cosmology. Our model involves four independent lognormal random variables: dark halo concentration c, disk spin lambdagal, disk mass fraction mgal, and stellar mass-to-light ratio UpsilonI. A simultaneous match of the VL and RL zero points with adiabatic contraction requires low-c halos, but this model has V2.2~1.8Vvir (where V2.2 and Vvir are the circular velocity at 2.2 disk scale lengths and the virial radius, respectively), which will be unable to match the luminosity function (LF). Similarly models without adiabatic contraction but standard c also predict high values of V2.2/Vvir. Models in which disk formation induces an expansion rather than the commonly assumed contraction of the dark matter halos have V2.2~1.2Vvir, which allows a simultaneous fit of the LF. This may result from nonspherical, clumpy gas accretion, where dynamical friction transfers energy from the gas to the dark matter. This model requires low lambdagal and mgal values, contrary to naive expectations. However, the low lambdagal is consistent with the notion that disk galaxies predominantly survive in halos with a quiet merger history, while a low mgal is also indicated by galaxy-galaxy lensing. The smaller than expected scatter in the RL relation and the lack of correlation between the residuals of the VL and RL relations, respectively, imply that the scatter in lambdagal and in c needs to be smaller than predicted for LambdaCDM halos, again consistent with the idea that disk galaxies preferentially reside in halos with a quiet merger history.