Testing the limits of alternative cosmologies and chasing the dream data set

The universe is accelerating, yet we know very little about the mysterious energy making it behave this way. Scientists put forward model after model, proposing arrays of cosmological fluids capable of causing acceleration. These models are fitted to data, and results are compared against each other. But how good is the current data at telling models apart? And what makes a better data set? The first aim of this thesis is to investigate what else might be hiding in the existing supernova data. This is done by generating and fitting increasingly strange, parameter-rich cosmological models. The second aim is to define such boundaries for the supernova data (sample size and uncertainty) which, if met, will allow scientists to rule out some of the more extreme contenders. This thesis explores the limits of the current supernova data, as well as its future potential. For the first part I found that, as far as the currently available data is concerned, our underlying cosmology may be very wild indeed. And as for the perfect data set? That remains to be seen