Top left: Photometric (black curves and thin faint curves, corresponding to different realizations of the redshift distribution; described in Sect. 3.1.1) and spectroscopic (coloured curves) tomographic redshift distributions utilized in this work (described in Sect. 3) for a stage III survey. Top right: Galaxy bias function (blue curve; Eq. (2.18)) used to define the true galaxy bias, b_g, for each tomographic redshift sample. Bottom left: Magnification bias functions, F_m(z) (Eq. (2.6)), used to calculate lensing contributions to galaxy number count kernels (Sect. 2.1), shown on a log-linear axis with transitions at ±1. The faint-end slope of the luminosity function is estimated via the fitting formulae of Joachimi & Bridle (2010), taking the r-band limiting magnitude r_lim = 24.5 for the photometric (blue curve) and r_lim = 20.0 for the spectroscopic (dashed orange curve) synthetic galaxy samples. Bottom right: Intrinsic alignment power spectrum prefactor function F_IA(z) (blue curve; Eq. (2.16)) used to approximate matter-intrinsic and intrinsic-intrinsic power spectra for the computation of intrinsic alignment contributions to the shear and GGL correlations. The galaxy bias b_g(z) and magnification bias F_m(z) functions are evaluated for each tomographic sample i at the mean redshift ⟨z⟩_i of the sample, denoted by black (photometric) or coloured (spectroscopic) points atop each curve. The intrinsic alignment prefactor F_IA(z) also displays these points but is evaluated at the mean of the relevant kernel product n_i(χ)n_j(χ), or n_i(χ)q_j(χ), for a correlation between samples i and j.