Non-local galaxy bias and the Ising model

Proper interpretation of the data from future dense galaxy surveys (such as Euclid and WFIRST) requires characterization of the bias between the dark matter and galaxy power spectra. We show that at small scales (high k) this relationship is not described by a simple linear bias parameter, and the bias on these scales in fact exhibits significant non-locality. As a result, we present a method for characterizing this k-dependent, non-local bias by fitting the Ising bias model to the data smoothed on multiple scales.