Can cheap approximations reduce noise in galaxy survey statistics?

Measuring the large-scale structure of the universe requires precise estimates of how uncertain our measurements are—a calculation that normally demands hundreds of expensive simulations. This work pairs each simulation with a cheap Zeldovich approximation sharing the same starting conditions, then uses the known statistical properties of the Zeldovich field to cancel out noise. The result: covariance matrices converge 2–3 times faster, with dramatic improvements on the largest scales where accuracy matters most for cosmology.