Title:Detecting Relativistic Doppler in Galaxy Clustering with Tailored Galaxy Samples

Abstract:We present a method to obtain a high-significance detection of relativistic effects on cosmological scales. Measurements of such effects would be instrumental for our understanding of the Universe, as they would provide a further confirmation of the validity of general relativity as the correct description of the gravitational interaction, in a regime very far from that of strong gravity, where it has been tested to exquisite accuracy. Despite its relevance, the detection of relativistic effects has hitherto eluded us, mainly because they are stronger on the largest cosmic scales, plagued by cosmic variance. Our work focuses on the cosmological probe of galaxy clustering, describing the excess probability of finding pairs of galaxies at a given separation due to them being part of the same underlying cosmic large-scale structure. We focus on the two-point correlation function of the distribution of galaxies in Fourier space -- the power spectrum -- where relativistic effects appear as an imaginary contribution to the real power spectrum. By carefully tailoring cuts in magnitude/luminosity, we are able to obtain two samples (bright and faint) of the same galaxy population, whose cross-correlation power spectrum allows for a detection of the relativistic contribution. In particular, we optimise the definition of the samples to maximise the detection significance of the relativistic Doppler term for both a low-$z$ Bright Galaxy Sample and a high-$z$ H$\alpha$ emission line galaxy population.