
We investigate the potential of weak lensing by voids to test for deviations
from General Relativity. We calculate the expected lensing signal of a scalar
field with derivative couplings, finding that it has the potential to boost the
tangential shear both within and outside the void radius. We use voids traced
by Luminous Red Galaxies in SDSS to demonstrate the methodology of testing
these predictions. We find that the void central density parameter, as inferred
from the lensing signal, can shift from its GR value by up to 20% in some
galileon gravity models. Since this parameter can be estimated independently
using the galaxy tracer profiles of voids, our method provides a consistency
check of the gravity theory. Although galileon gravity is now disfavoured as a
source of cosmic acceleration by other datasets, the methods we demonstrate
here can be used to test for more general fifth force effects with upcoming
void lensing data.

We measure weak lensing mass profiles of voids from a volumelimited sample
of SDSS Luminous Red Galaxies (LRGs). We find voids using an algorithm designed
to maximize the lensing signal by dividing the survey volume into 2D slices,
and then finding holes in this 2D distribution of LRGs. We perform a stacked
shear measurement on about 20,000 voids with radii between 1555 Mpc/h and
redshifts between 0.160.37. We measure the characteristic radial shear signal
of voids with a signaltonoise of 7. The mass profile corresponds to a
fractional underdensity of about 0.4 inside the void radius and a slow
approach to the mean density indicating a partially compensated void structure.
We compare our measured shape and amplitude with the predictions of Krause et
al 2013. Voids in the galaxy distribution have been extensively modeled using
simulations and measured in the SDSS. We discuss how the addition of void mass
profiles can enable studies of galaxy formation and cosmology.

We present a novel method to detect the effects of dynamical friction in
observed galaxy clusters. Following accretion into clusters, massive satellite
galaxies will backsplash to systematically smaller radii than less massive
satellites, an effect that may be detected by stacking the number density
profiles of galaxies around clusters. We show that this effect may be
understood using a simple toy model which reproduces the trends with halo
properties observed in simulations. We search for this effect using SDSS
redMaPPer clusters with richness 10<lambda<20, and find that bright (M_i<21.5)
satellites have smaller splashback radii than fainter (M_i>20) satellites at
99% confidence.

When dark matter halos are accreted by massive host clusters, strong
gravitational tidal forces begin stripping mass from the accreted subhalos.
This stripping eventually removes all mass beyond a subhalo's tidal radius, but
the unbound mass remains in the vicinity of the satellite for at least a
dynamical time t_dyn. The Nbody subhalo study of Chamberlain et al. verified
this picture and pointed out a useful observational consequence: measurements
of subhalo correlations beyond the tidal radius are sensitive to the infall
time, t_infall, of the subhalo onto its host. We perform this crosscorrelation
measurement using ~ 160,000 red satellite galaxies in SDSS redMaPPer clusters
and find evidence that subhalo correlations do persist well beyond the tidal
radius, suggesting that many of the observed satellites fell into their current
host less than a dynamical time ago, t_infall < t_dyn. Combined with estimated
dynamical times t_dyn ~ 35 Gyr and SED fitting results for the time at which
satellites stopped forming stars, t_quench ~ 6 Gyr, we infer that for a
significant fraction of the satellites, star formation quenched before those
satellites entered their current hosts. The result holds for red satellites
over a large range of clustercentric distances 0.1  0.6 Mpc/h. We discuss the
implications of this result for models of galaxy formation.

We present cosmological constraints from the Dark Energy Survey (DES) using a
combined analysis of angular clustering of red galaxies and their
crosscorrelation with weak gravitational lensing of background galaxies. We
use a 139 square degree contiguous patch of DES data from the Science
Verification (SV) period of observations. Using large scale measurements, we
constrain the matter density of the Universe as Omega_m = 0.31 +/ 0.09 and the
clustering amplitude of the matter power spectrum as sigma_8 = 0.74 +/ 0.13
after marginalizing over seven nuisance parameters and three additional
cosmological parameters. This translates into S_8 = sigma_8(Omega_m/0.3)^{0.16}
= 0.74 +/ 0.12 for our fiducial lens redshift bin at 0.35 <z< 0.5, while S_8 =
0.78 +/ 0.09 using two bins over the range 0.2 <z< 0.5. We study the
robustness of the results under changes in the data vectors, modelling and
systematics treatment, including photometric redshift and shear calibration
uncertainties, and find consistency in the derived cosmological parameters. We
show that our results are consistent with previous cosmological analyses from
DES and other data sets and conclude with a joint analysis of DES angular
clustering and galaxygalaxy lensing with Planck CMB data, Baryon Accoustic
Oscillations and Supernova type Ia measurements.

Lensing measurements of the shapes of dark matter halos can provide tests of
gravity theories and possible dark matter interactions. We measure the
quadrupole weak lensing signal from the elliptical halos of 70,000 SDSS
Luminous Red Galaxies. We use a new estimator that nulls the spherical halo
lensing signal, isolating the shear due to anisotropy in the dark matter
distribution. One of the two Cartesian components of our estimator is
insensitive to the primary systematic, a spurious alignment of lens and source
ellipticities, allowing us to make robust measurements of halo ellipticity. Our
bestfit value for the ellipticity of the surface mass density is $0.24 \pm
0.06$, which translates to an axis ratio of 0.78. We rule out the hypothesis of
no ellipticity at the $4\sigma$ confidence level, and ellipticity < 0.12 (axis
ratio > 0.89) at the $2\sigma$ level. We discuss how our measurements of halo
ellipticity are revised to higher values using estimates of the misalignment of
mass and light from simulations. Finally, we apply the same techniques to a
smaller sample of redMaPPer galaxy clusters and obtain a $3\sigma$ measurement
of cluster ellipticity. We discuss how the improved signal to noise properties
of our estimator can enable studies of halo shapes for different galaxy
populations with upcoming surveys.

We search for the lensing signal of massive filaments between 135,000 pairs
of Luminous Red Galaxies (LRGs) from the Sloan Digital Sky Survey. We develop a
new estimator that cleanly removes the much larger shear signal of the
neighboring LRG halos, relying only on the assumption of spherical symmetry. We
consider two models: a thick filament model constructed from raytracing
simulations for $\Lambda$CDM model, and a thin filament model which models the
filament by a string of halos along the line connecting the two LRGs. We show
that the filament lensing signal is in nice agreement with the thick simulation
filament, while strongly disfavoring the thin model. The magnitude of the
lensing shear due to the filament is below $10^{4}$. Employing the likelihood
ratio test, we find a 4.5$\sigma$ significance for the detection of the
filament lensing signal, corresponding to a null hypothesis fluctuation
probability of $3 \times 10^{6}$. We also carried out several null tests to
verify that the residual shear signal from neighboring LRGs and other shear
systematics are minimized.

Using a void catalog from the SDSS survey, we present the first measurements
of void clustering and the corresponding void bias. Over the range 30200 Mpc/h
the void autocorrelation is detected at 5sigma significance for voids of
radius 1520 Mpc/h. We also measure the voidgalaxy crosscorrelation at higher
signaltonoise and compare the inferred void bias with the autocorrelation
results. Void bias is constant with scale for voids of a given size, but its
value falls from 5.6 +/ 1.0 to below zero as the void radius increases from 15
to 30 Mpc/h. The comparison of our measurements with carefully matched galaxy
mock catalogs, with no free parameters related to the voids, shows that model
predictions can be reliably made for void correlations. We study the dependence
of void bias on tracer density and void size with a view to future
applications. In combination with our previous lensing measurements of void
mass profiles, these clustering measurements provide another step towards using
voids as cosmological tracers.

We investigate the behavior of the fifth force in voids in chameleon models
using the spherical collapse method. Contrary to Newtonian gravity, we find the
fifth force is repulsive in voids. The strength of the fifth force depends on
the density inside and outside the void region as well as its radius. It can be
many times larger than the Newtonian force and their ratio is in principle
unbound. This is very different from the case in halos, where the fifth force
is no more than 1/3 of gravity. The evolution of voids is governed by the
Newtonian gravity, the effective dark energy force and the fifth force. While
the first two forces are common in both LCDM and chameleon universes, the fifth
force is unique to the latter. Driven by the outwardpointing fifth force,
individual voids in chameleon models expand faster and grow larger than in a
LCDM universe. The expansion velocity of the void shell can be 20% to 30%
larger for voids of a few Mpc/h in radius, while their sizes can be larger by
~10%. These differences are smaller for larger voids of the same density. We
compare void statistics using excursion set theory; for voids of the same size,
their number density is larger in chameleon models. The fractional difference
increases with void size. The chance of having voids of radius ~25 Mpc/h can be
2.5 times larger. This difference is about 10 times larger than that in the
halo mass function. We find strong environmental dependence of void properties
in chameleon models. The differences in size and expansion velocity with GR are
both larger for small voids in high density regions. In general, the difference
between chameleon models and LCDM in void properties (size, expansion velocity
and distribution function) are larger than the corresponding quantities for
halos. This suggests that voids might be better candidates than halos for
testing gravity.

We offer predictions of symmetron modified gravity in the neighborhood of
realistic dark matter halos. The predictions for the fifth force are obtained
by solving the nonlinear symmetron equation of motion in the spherical NFW
approximation. In addition, we compare the three major known screening
mechanisms: Vainshtein, Chameleon, and Symmetron around such dark matter
sources, emphasizing the significant differences between them and highlighting
observational tests which exploit these differences. Finally, we demonstrate
the host halo environmental screening effect ("blanket screening") on smaller
satellite halos by solving for the modified forces around a density profile
which is the sum of satellite and approximate host components.