An improved cosmological model fitting of Planck data with a dark energy spike

Abstract

The $\Lambda$ cold dark matter ($\Lambda\textrm{CDM}$) model is currently known as the simplest cosmology model that best describes observations with minimal number of parameters. Here we introduce a cosmology model that is preferred over the conventional $\Lambda\textrm{CDM}$ one by constructing dark energy as the sum of the cosmological constant $\Lambda$ and the additional fluid that is designed to have an extremely short transient spike in energy density during the radiation-matter equality era and the early scaling behavior with radiation and matter densities. The density parameter of the additional fluid is defined as a Gaussian function plus a constant in logarithmic scale-factor space. Searching for the best-fit cosmological parameters in the presence of such a dark energy spike gives a far smaller chi-square value by about five times the number of additional parameters introduced and narrower constraints on matter density and Hubble constant compared with the best-fit $\Lambda\textrm{CDM}$ model. The significant improvement in reducing chi-square mainly comes from the better fitting of Planck temperature power spectrum around the third ($\ell \approx 800$) and sixth ($\ell \approx 1800$) acoustic peaks. The likelihood ratio test and the Akaike information criterion suggest that the model of dark energy spike is strongly favored by the current cosmological observations over the conventional $\Lambda\textrm{CDM}$ model. However, based on the Bayesian information criterion which penalizes models with more parameters, the strong evidence supporting the presence of dark energy spike disappears. Our result emphasizes that the alternative cosmological parameter estimation with even better fitting of the same observational data is allowed in the Einstein's gravity.