Millilensing statistics for BH clusters and Gaussian subhaloes

Abstract

The aim of this TFM is to probe candidates to Dark Matter through their effect on the probability density function of the magnification induced by gravitational lensing, P DF(µ). This is done by simulating the millilensing effect using the inverse ray tracing algorithm in both Python3 and FORTRAN95. Furthermore, we have implemented the nearest neighbours algorithm in the codes and performed a comparison of the computational efficiency between Python3 and FORTRAN95 scripts. To cross-check our numerical models we compare with the theoretical predictions for the sparse case. Our first candidates to dark matter are primordial black holes, which become an interesting possibility after LIGO-VIRGO observations. In this TFM we are considering them grouped in clusters of variable compactness, instead of following a random uniform distribution. Furthermore, we study whether at large scales a compact cluster behaves as a pseudo-particle being indistinguishable from a single black hole with the same mass of the cluster. Our second candidates under study are DM subhaloes. The existence of these subhaloes is predicted by CDM models of structure formation. We analyze whether the gravitational lensing magnification induced by DM Gaussian subhaloes with different compactness is consistent with real observations. A Bayesian analysis based on compactness and magnification shows that, according to observations, clustering makes less probable the existence of PBH’s and that a large compactness of the subhaloes should result in millilensing magnification larger than observed.