The effect of baryonic physics on gravitational lensing. The red and magenta dashed lines show the sample variance limit for a 10% band in ell assuming fsky=0.1 (red) and fsky=1 (magenta). The dotted lines include the shot noise for 25, 50 and 100 galaxies per square arcminute. The blue line shows a model for the effects of baryonic cooling on the power spectrum.

The "swirling" pattern of shear which would be a pure B-mode signal. This is not generated by lensing at lowest order.

A convergence field, 3 degrees on a side, with the 32 most massive clusters in the field circled. Note that the most massive clusters don't correspond 1-1 with the largest convergence peaks. Shear selection is not equivalent to mass selection.

A slice, 100Mpc/h on a side, through an N-body simulation, showing that clusters form at the intersection of a beaded filamentary network. The region around clusters is typically more dense than an average region of the universe, and multiple smaller halos (associated with the cluster) can be seen.

A convergence map, 0.3 degrees on a side, showing the effects of projection. The left panel shows the total convergence. The peak is actually two much lower mass peaks which happen to line up along the line-of-sight. The middle panel shows the closer of the two and the right panel the further of the two halos.

The tangential pattern of shear which would occur around an overdensity. This is a pure E-mode signal from an isolated, spherically symmetric mass.

The angular power spectrum of the convergence for a currently popular Lambda CDM cosmology. The linear and non-linear power spectra are plotted along with the shot noise power for 25, 50 and 100 galaxies per square arcminute.

Lensing tomography: the angular power spectrum matrix is plotted for 2-bin tomography. There are two source bins (one at z~1 and one at z~1.5). The pairs of galaxies which make up the 2-point function can either come both from the first bin (11), both from the second (22) or one from each (12). The signal in the two bins is highly correlated, r>0.9 in this case. ~

A shear map, 3 degrees on a side, coarsely sampled. This is the first of 4 panels illustrating aperture masses. Note the tangential signals around mass peaks in the map.

The variance of the aperture mass, computed from the shear map of the previous plot, as a function of scale. The solid line is the E-mode (computed using gamma_+) while the dashed line is the B-mode (computed using gamma_x). Note that due to lens-lens coupling and other higher order effects the B-mode from lensing is not zero, just small.

The M_ap field computed from the shear field (above). Note the peaks in the M_ap corresponding to clusters and groups of clusters at the positions where tangential shear is visible in the shear map.

The M_perp field (i.e. M_ap with gamma_x so as to measure the B-mode signal) computed from the shear field of the earlier plot. While there is structure in the map, it is very small compared to the E-mode signal (previous picture, same color scale).

A comparison of P(k) derived from N-body simulations of a LCDM model with the analytic fitting function of the Virgo consortium. The agreement is quite good, except at high-k. The fluctuations at low-k are due to sample variance ... two different runs with the same P(k) but different initial phases are shown to illustrate the sampling error. This illustrates that independent codes to generate initial conditions, independent N-body and analysis codes produce very similar power spectra -- within the quoted limits of the Virgo consortium fitting function up to k~10h/Mpc.

A convergence map, linear color scale running from -0.1 to 0.1, with the shear field overplotted. ~

The effect of source density on the ability to reconstruct images of the mass density. The left panel shows a convergence field from a simulation, 3 degrees on a side with a linear color scale running from -0.1 to 0.1. The middle panel shows the same field, with 50 galaxies per square arcminute sampling (optimistic from the ground) and the right panel 200 galaxies per square arcminute (a deep space image).

Return to Martin's home page.

Back to data archive.

*Last modified Thu Jun 24 13:03:14 2004*