The Multi-Tracer Optimal Estimator applied to VIPERS

We use mock galaxy data from the VIMOS Public Extragalactic Redshift Survey (VIPERS) to test the performance of the Multi-Tracer Optimal Estimator (MTOE) of Abramo et al. as a tool to measure the monopoles of the power spectra of multiple tracers of the largescale structure, P-alpha((0))(k). We show that MTOE provides more accurate measurements than the standard technique of Feldman, Kaiser&Peacock (FKP), independently of the tracer-selection strategy adopted, on both small and large scales. The largest improvements on individual P-alpha((0))(k) are obtained on small scales, using a colour-magnitude selection, due to MTOE being naturally better equipped to deal with shot noise: we report an average error reduction with respect to FKP of similar to 30 per cent at 0.3 < k[h Mpc(-1)] < 0.5, with improvements exceeding 40-50 per cent for some tracers. On large scales (k[h Mpc(-1)] less than or similar to 0.1), the gain in accuracy resulting from cosmic-variance cancellation is similar to 10 per cent for the ratios of P-alpha((0))(k). We have carried out a Markov chain Monte Carlo analysis to determine the impact of these gains on several quantities derived from P-alpha((0))(k). If we push that the estimated power spectra are themselves the measurement to scales 0.3 < k[h Mpc(-1)] < 0.5, the average improvements are similar to 30 per cent for the amplitudes of the monopoles, similar to 75 per cent for the monopole ratios, and similar to 20 per cent for the linear galaxy biases. Our results highlight the potential of MTOE to shed light upon the physics that operate both on large and small cosmological scales. The effect of MTOE on cosmological constraints using VIPERS data will be addressed in a separate paper.