Cross-correlating spectroscopic and photometric galaxy surveys

Abstract

In this thesis we study constraining cosmology when combining spectroscopic and photometric galaxy survey. The photometric survey measures galaxy shape distortions from Weak Lensing (WL), while high precision redshift information makes spectroscopic surveys ideal for redshift space distortions (RSD). The combined analysis is performed entirely in angularcorrelation functions, which simplifies the joined analysis, in particular the inclusion of covariance between then.The first chapter introduce a novel algorithm for efficiently calculating the cross-correlations of multiple tracers (i.e. galaxy types/luminosities) and including WL in narrow redshift bin cross-correlations. Estimating the angular-correlations function is in particular demanding since the number for cross-correlations increase O(n 2 ) with n being the number of redshift bins.Later the chapter study the effect of Limber approximation and RSD on the modeling of auto-and cross-correlations. For thin redshift bins, the Limber approximation completely breaks down and does not allow crosscorrelations between redshift bins. Decreasing the bin width increases the amplitude of the galaxy correlations and the effect of RSD, which will benefit the cosmological constraints. One interesting trend is the baryon acoustic oscillation (BAO) contribution in the cross-correlations of redshift bins. The redshift separations between two bins reduce small-scale clustering, hence increasing the BAO contrast. We also study the signal-to-noise of different cross-correlations.The second chapter forecast the constraints on the cosmic expansion and growth history, using two fiducial 14000 sq deg. spectroscopic and photometric galaxy surveys. Overlapping surveys (same sky) has improved constraints from additional cross-correlations and sample variance cancellations (covariance in multiple tracers). We study first separate how redshift bin width, RSD, BAO and WL affect the forecast. We find gains equivalent to 30% larger areas when using overlapping surveys. Last, we discuss the origen v of this moderate gain in the context of existing literature. Different groups reports either none or high benefits for overlapping galaxy surveys. We suggest the covariance between surveys and different same-sky definitions (i.e. different observables) can explain the differences.Galaxy bias relate the galaxy overdensities to the underlying matter fluctuations, and the uncertainty in galaxy bias strongly affects the forecast. We therefore investigate in detail how cross-correlations, RSD, BAO and WL affects constraints on galaxy bias. Overlapping surveys in particular increase constraint on the bias from the photometric sample. Last section quantify the benefit of priors and the effect of bias stochasticity. The impact of uncertainties in bias stochasticity is less for overlapping surveys. ResumenEn esta tesis se estudia el acotamiento en los paraámetros cosmología al combinar observaciones en catálogos de galaxias espectroscópicas y fotométricos. Los catálogos fotométricos miden las d...