The MIP ensemble simulation: local ensemble statistics in the Cosmic Web

We present a new technique that allows us to compute ensemble statistics on a local basis, directly relating halo properties to their local environment. This is achieved by the use of a correlated ensemble in which the large-scale structure (LSS) is common to all realizations while having each an independent halo population. The correlated ensemble can be stacked, effectively increasing the halo number density by an arbitrary factor, thus breaking the fundamental limit in the halo number density given by the halo mass function. This technique allows us to compute local ensemble statistics of the matter/halo distribution at any position in the simulation box, while removing the intrinsic stochasticity in the halo formation process and directly relating halo properties to their environment. We introduce the Multum In Parvo correlated ensemble simulation consisting of 220 realizations on a 32 h(-1) Mpc box with 256(3) particles each. This is equivalent in terms of effective volume and number of particles to a box of similar to 193 h(-1) Mpc of side with similar to 1540(3) particles containing similar to 5 x 10(6) haloes with a minimum mass of 3.25 x 10(9) h(-1) M-circle dot. The potential of the technique presented here is illustrated by computing the local ensemble statistics of the halo ellipticity and halo shape-LSS alignment. We show that, while there are general trends in the ellipticity and alignment of haloes with their LSS, there are also significant spatial variations which has important implications for observational studies of galaxy shape and alignment.