Day 1118

Friday.

1607.01782
Approximate Bayesian Computation in large scale structure: constraining the galaxy-halo connection
Hahn, et al

The standard approaches to Bayesian parameter inference in LSS assume a Gaussian functional from (chi2 form) for the likelihood.  They are also typically restricted to measurements such as the 2pcf.  Likelihood free inferences such as Approximate Bayesian Computation (ABC) make inference possible without assuming any functional form for the likelihood, thereby relaxing the assumptions and restrictions of the standard approach.  Instead it relies on a forward generative model of the data and a metric for measuring the distance between the model and data.  In this work, demonstrate that ABC is feasible for LSS parameter inference by using it to constraining parameters of the HOD model for populating DM haloes with galaxies.  Using specific implementation of ABC supplemented with Population Monte Carlo importance sampling, a generative forward model using HOD, and a distance metric based on galaxy number density, 2pcf, and galaxy group multiplicity function, constrain the HOD parameters of mock observation generated rom selected "true" HOD parameters.  The parameter constraints obtained from ABC are consistent with the "true" HOD parameters, demonstrating that ABC can be reliably used for parameter inference in LSS.  Furthermore, compare the ABC constraints to constraints obtained using a pseudo-likelihood function of Gaussian form with MCMC and find consistent HOD parameter constraints.  Ultimately the results suggest that ABC can and should be applied in parameter inference of LSS analyses.


1607.02056
Shear Nulling after PSF Gaussianization: moment-based weak lensing measurements with sub percent noise bias
Herbonnet, Buddendiek, Kuijken

Current optical imaging surveys for cosmology are covering large areas of sky.  To exploit the statistical power of these surveys for WL measurements requires shape measurement methods with sub percent systematic errors.  Introduce a new WL shear measurement algorithm, SNAPG, designed to avoid the noise biases that affect most other methods.  SNAPG operates on images that have been convolved with a kernel that renders the PSF a circular Gaussian, and uses weighted second moments of the sources.  The response of such second moments to a shear of the pre-seeing galaxy image can be predicted analytically, allowing to construct a shear nulling scheme that finds the shear parameters for which the observed galaxies are consistent with an unsheared, isotropically oriented population of sources.  The inverse of this nulling shear is then an estimate of the gravitational lensing shear.  Identify the uncertainty of the estimated center of each galaxy as the source of noise bias, and incorporate an approximate estimate of the centroid covariance into the scheme.  Test the method on extensive suits of simulated galaxies of increasing complexity, and find that it is capable of shear measurement with multiplicative bias below 0.5%.