We apply the Karhunen-Lo\'eve methods to cosmic microwave background (CMB) data sets, and show that we can recover the input cosmology and obtain the marginalized likelihoods in $\Lambda$ cold dark matter cosmologies in under a minute, much faster than Markov chain Monte Carlo methods. This is achieved by forming a linear combination of the power spectra at each multipole $l$, and solving a system of simultaneous equations such that the Fisher matrix is locally unchanged. Instead of carrying out a full likelihood evaluation over the whole parameter space, we need evaluate the likelihood only for the parameter of interest, with the data compression effectively marginalizing over all other parameters. The weighting vectors contain insight about the physical effects of the parameters on the CMB anisotropy power spectrum $C_l$. The shape and amplitude of these vectors give an intuitive feel for the physics of the CMB, the sensitivity of the observed spectrum to cosmological parameters, and the relative sensitivity of different experiments to cosmological parameters. We test this method on exact theory $C_l$ as well as on a Wilkinson Microwave Anisotropy Probe (WMAP)-like CMB data set generated from a random realization of a fiducial cosmology, comparing the compression results to those from a full likelihood analysis using CosmoMC. After showing that the method works, we apply it to the temperature power spectrum from the WMAP seven-year data release, and discuss the successes and limitations of our method as applied to a real data set.

We determine the forecast errors on the absolute neutrino mass scale and the equation of state of dark energy by combining synthetic data from the Dark Energy Survey (DES) and the Cosmic Microwave Background (CMB) Planck surveyor. We use angular clustering of galaxies for DES in 7 redshift shells up to $z\sim 1.7$ including cross-correlations between different redshift shells. We study models with massless and massive neutrinos and three different dark energy models: $\Lambda$CDM ($w=-1$), wCDM (constant $w$), and waCDM (evolving equation of state parameter $w(a)=w_0 + w_{a}(1-a)$). We include the impact of uncertainties in modeling galaxy bias using a constant and a redshift-evolving bias model. For the $\Lambda$CDM model we obtain an upper limit for the sum of neutrino masses from DES+Planck of $\Sigma m_\nu < 0.08$ eV (95\% C.L.) for a fiducial mass of $\Sigma m_\nu = 0.047$ eV, with a 1$\sigma$ error of 0.02 eV, assuming perfect knowledge of galaxy bias. For the wCDM model the limit is $\Sigma m_\nu < 0.10 $ eV. For a wCDM model where galaxy bias evolves with redshift, the upper limit on the sum of neutrino masses increases to 0.19 eV. DES will be able to place competitive upper limits on the sum of neutrino masses of 0.1-0.2 eV and could therefore strongly constrain the inverted mass hierarchy of neutrinos. In a wCDM model the 1$\sigma$ error on constant $w$ is $\Delta w= 0.03$ from DES galaxy clustering and Planck. Allowing $\Sigma m_\nu$ as a free parameter increases the error on $w$ by a factor of 2, with $\Delta w=0.06$. In a waCDM model, in which the dark energy equation of state varies with time, the errors are $\Delta w_0 = 0.2$ and $\Delta w_a = 0.42$. Including neutrinos and redshift dependent galaxy bias increases the errors to $\Delta w_0 = 0.35$ and $\Delta w_a = 0.89$.

Possible explanations of the observed accelerated expansion of the Universe are the introduction of a dark energy component or the modifications of gravity at large distances. A particular difference between these scenarios is the dynamics of the growth of structures. The redshift distribution of galaxy clusters will probe this growth of structures with large precision. Here we will investigate how proposed galaxy cluster surveys will allow one to distinguish the modified gravity scenarios from dark energy models. We find that cluster counts can distinguish the Dvali-Gabadadze-Porrati model from a dark energy model, which has the same background evolution, as long as the amplitude of the primordial power spectrum is constrained by a CMB experiment like Planck. In order to achieve this, only a couple of hundred clusters in bins of width Delta-z = 0.1 are required. This should be easily achievable with forthcoming Sunyaev-Zel'dovich cluster counts, such as the South Pole Telescope in conjunction with the Dark Energy Survey.