Shear, B1608 and long tails.
Shear
With our convergence recipe, it is easy to also predict the external shear along a line of sight
Fig 1 Shear and kappa reconstructions for a randomly selected line of sight. Black star shows the ray traced result for this line of sight, and blue star shows the reconstruction given perfect knowledge of halo masses and redshifts (for all halos out to 5 arcmins and down to i=26)
The Reconstructions have asymmetric long tails. The prescription for going from stellar mass -> halo mass -> kappa is very assymetric: there is a long tail out to high kappa. This raises several issues.
-
how do we correctly do the zero correction - the long tails grossly skew the mean.
-
The long tails are unphysical - the probability in the tails is much higher than there should be, given the global P(kappa) of the universe
How should we deal with these questions? An envelope function? Form the joint distribution P(kappa_rec,kappa_ray) and marginalize over P(kappa_rec)?
Fig 2 Example reconstructions for a random line of sight. Notice the tails. The same zero point is used for all of them: the zero point needed to make mean kappa_reconstruct = 0 given perfect knowledge of halo mass. This is not necessarily fair.
Taking Zach's data for the B1608 line of sight, we can reconstruct this sight line:
Fig 3 Reconstruction for B1608. Note that without solving the zeropoint issue we can't be sure that the x scale is correct. I've used the correction from 2000 perfect mass reconstructions of 1 arcmin apperture down to i=26.
Fig 4 Reconstruction for B1608. Shear reconstruction too. Orientation of phi is east through north.