On mitigation of the uncertainty in non-linear matter clustering for cosmic shear tomography

We present a new method that deals with the uncertainty in matter clustering in cosmic shear power spectrum analysis that arises mainly due to poorly understood non-linear baryonic processes on small scales. We show that the majority of information about new physics contained in the shear power comes from these small scales. Removing these non-linear scales from a cosmic shear analysis results in 50 per cent cut in the accuracy of measurements of dark energy parameters, marginalizing over all other parameters. In this paper we propose a method to recover the information on small scales by allowing cosmic shear surveys to measure the non-linear matter power spectrum themselves and marginalize over all possible power spectra using path integrals. Information is still recoverable in these non-linear regimes from the geometric part of weak lensing. In this self-calibration regime we recover 90 per cent of the information on dark energy. Including an informative prior, we find that the nonlinear matter power spectrum needs to be accurately known to 1 per cent down to k = 50 h(-1) Mpc, or a scale of 120 kpc, to recover 99 per cent of the dark energy information. This presents a significant theoretical challenge to understand baryonic effects on the scale of galaxy haloes.