Dark Energy Constraints and Joint Cosmological Inference from Mutually Inconsistent Observations

Abstract

Recent results regarding dark energy are mutually inconsistent under the ΛCDM cosmological model, hinting at the possibility of undiscovered physics. However, the currently accepted cosmological parameters come from a joint inference between observational data sets, a process that is formally invalid for inconsistent observations. We will show that many problems can arise when using joint inference on disagreeing observations such as significantly underestimated margins of error, high dependencies on priors, and sensitivities to boundary constraints. Because we do not know if these inconsistencies arise due to errors in observation, poor statistical techniques, or an improper model, it is difficult to fix these problems. Each scenario in which the analysis method breaks is discussed alternative resampling techniques are explored, demonstrating that ΛCDM should not yet be rejected based on current data.