Examining a new form of non-standard dark matter using DESI DR2 data

TL;DR

This work addresses potential deviations from cold dark matter by introducing a non-standard DM with a continuous equation of state $w_{dm}=w2 a^2$, designed to approach CDM at early times. It jointly constrains this NSDM with three DE parameterizations (Lambda, constant w, and CPL) using Planck2018 CMB distance priors, DESI DR2 BAO, and SN Ia data through MCMC, comparing to LambdaCDM via chi-square differences. The results consistently favor negative $w2$, particularly in the ww2_DM variant, and show NSDM can substantially reduce the probability of violating the null energy condition, with several data combinations indicating a better fit than LambdaCDM on the order of 1.7–3.6σ. The CPL case weakens the DM non-cold signal, implying sensitivity to the chosen DE sector, and overall NSDM models can match or surpass LambdaCDM in explanatory power for current data.

Abstract

In this work, we propose a non-standard dark matter (NSDM) model in which the equation of state (EoS) of dark matter (DM) is parameterized as $w_{\rm dm} = w_2 a^2$, and this DM model is motivated by the idea that DM must become cold dark matter (CDM) in the neighborhood of the scale factor $a = 0$, which implies that both the EoS of DM, $w_{\rm dm}$, and its derivative with respect to the scale factor, ${\rm d}w_{\rm dm}/{\rm d}a$, vanish at $a = 0$. By incorporating the latest cosmological datasets -- including the Planck2018 Cosmic Microwave Background (CMB) distance priors, the Baryon Acoustic Oscillation measurements from the Data Release 2 of the Dark Energy Spectroscopic Instrument (DESI), together with three independent Type Ia Supernova datasets, namely the Dark Energy Survey Year 5 (DESY5) compilation, the Union3 compilation, and the PantheonPlus sample -- we constrain the $Λw_2$DM, $ww_2$DM, and $w_0w_aw_2$DM models, which are constructed by replacing CDM with NSDM in the $Λ$CDM, $w$CDM, and $w_0w_a$CDM models, respectively. We find that there is a preference for a negative DM EoS at more than the $3σ$ confidence level for the data combinations CMB+DESI+Union3 and CMB+DESI+DESY5. Moreover, for all data combinations, replacing CDM with NSDM in the $w$CDM and $w_0w_a$CDM models significantly reduces the probability of violating the null energy condition. Furthermore, both $ww_2$DM and $w_0w_aw_2$DM are favored over $Λ$CDM with a significance comparable to that of the $w_0w_a$CDM model.

Figures (5)

• Figure 1: The one-dimensional marginalized posterior distributions of $w_2$ for the $\Lambda w_2$DM model using CMB+DESI, CMB+DESI+PP, CMB+DESI+Union3, and CMB+DESI+DESY5 data combinations. The dashed line corresponds to $w_2=0$.
• Figure 2: The two-dimensional joint contours at $1\sigma$ and $2\sigma$ CL of parameters $w$ and $w_2$ (left panel) and one-dimensional marginalized posterior distributions of $w_2$ (right panel) for the $ww_2$DM model using CMB+DESI, CMB+DESI+PP, CMB+DESI+Union3, and CMB+DESI+DESY5 data combinations. The horizontal dashed line corresponds to $w=-1$, and two vertical dashed lines corresponds to $w_2=0$.
• Figure 3: The triangular plot of the fitting results for the $w_0w_aw_2$DM model using CMB+DESI, CMB+DESI+PP, CMB+DESI+Union3, and CMB+DESI+DESY5 data combinations. The dashed lines correspond to the three cases: $w_0=-1$, $w_a=0$, and $w_2=0$.
• Figure 4: The binned one-dimensional marginalized posterior distributions of $w$ for the $w$CDM and $ww_2$DM models under CMB+DESI (top left), CMB+DESI+PP (top right), CMB+DESI+Union3 (bottom left), and CMB+DESI+DESY5 (bottom right) data combinations. The dashed line corresponds to $w=0$.
• Figure 5: The binned one-dimensional marginalized posterior distributions of the crossing scale factor $a_{\mathrm{across}}$ for the $w_0w_a$CDM and $w_0w_aw_2$DM models under CMB+DESI (top left), CMB+DESI+PP (top right), CMB+DESI+Union3 (bottom left), and CMB+DESI+DESY5 (bottom right) data combinations. two dashed lines corresponds to $a_{\mathrm{across}}=0$ and $a_{\mathrm{across}}=1$, respectively.