LAUSR

We study reionization in two non-flat $\Lambda$CDM inflation models that best fit the Planck 2015 cosmic microwave background anisotropy observations, ignoring or in conjunction with baryon acoustic oscillation distance measurements. We implement a principal component analysis (PCA) to estimate the uncertainties in the reionization history from a joint quasar-CMB dataset. A thorough Markov Chain Monte Carlo analysis is done over the parameter space of PCA modes for both non-flat $\Lambda$CDM inflation models as well as the original Planck 2016 tilted, spatially-flat $\Lambda$CDM inflation model. Although both flat and non-flat models can closely match the low-redshift ($z\lesssim6$) observations, we notice a possible tension between high-redshift ($z\sim8$) Lyman-$\alpha$ emitter data and the non-flat models. This is solely due to the fact that the closed models have a relatively higher reionization optical depth compared to the flat one, which in turn demands more high-redshift ionizing sources and favors an extended reionization starting as early as $z\approx14$. We conclude that as opposed to flat-cosmology, for the non-flat cosmology models (i) the escape fraction needs steep redshift evolution and even unrealistically high values at some redshifts and (ii) most of the physical parameters require to have non-monotonic redshift evolution, especially apparent when Lyman-$\alpha$ emitter data is included in the analysis.