The variance and covariance of counts-in-cells probabilities

Abstract

Counts-in-cells (CIC) measurements contain a wealth of cosmological information, yet are seldom used to constrain theories. Although we can predict the shape of the distribution for a given cosmology, to fit a model to the observed CIC probabilities requires the covariance matrix - both the variance of counts in one probability bin and the covariance between counts in different bins. To date, there have been no general expressions for these variances. Here, we show that correlations of particular levels, or 'slices', of the density field determine the variance and covariance of CIC probabilities. We derive explicit formulae that accurately predict the variance and covariance among subvolumes of a simulated galaxy catalogue, opening the door to the use of CIC measurements for cosmological parameter estimation.