With the next generation of wide field galaxy surveys, both spectroscopic and photometric, we expect to achieve unprecedented constraints on the expansion history of the universe and the growth of structure. Maximizing the flow of information from these rich datasets to constraints on our physical models requires accurate characterization of systematic uncertainties. First, we present a method for estimation of covariance matrices of galaxy clustering measurements with spectroscopic surveys. We show that our method enables us to generate accurate galaxy mocks needed for BAO and RSD analyses on nonlinear scales. Then, we present the main challenges in extracting cosmological information from lensing measurements of deep imaging surveys. We show that employing novel techniques in estimation of the point spread function can keep this major systematic under control. Finally, we discuss various approaches for improvement of the photometric redshifts for the imaging surveys. We demonstrate how the precision and accuracy of photometric redshifts can be greatly enhanced if we take advantage of combining different datasets.