Exploring the Impact of Systematic Bias in Type Ia Supernova Cosmology Across Diverse Dark Energy Parametrizations

TL;DR

The paper assesses how instrumental and astrophysical SN-Ia systematics bias constraints on dark energy across four parametrizations, including the robust Generalised Scale Factor (GEN) form and the common CPL, JBP, and LOG models, using simulated Pantheon-Plus SN-Ia data with DESI BAO and compressed CMB anchors. Six systematics (calibration, intergalactic dust, luminosity evolution, stretch evolution, color-scatter, and density-mismatch) are modeled to quantify biases in the DE parameters $w_0$ and $w_a$; calibration offsets and progenitor evolution dominate the error budget. The results show JBP is most sensitive to systematics while GEN remains the least affected, with CPL/LOG showing intermediate sensitivity, highlighting how the choice of parametrization shapes inferred dark energy evolution. The findings stress sub-percent calibration precision and better astrophysical modelling as crucial for robust conclusions about dynamical dark energy, and warn that apparent time-variation signals could arise from unaccounted SN-Ia systematics. Future work should emphasize cross-survey calibration, multi-wavelength SN analyses, and joint SN–BAO–CMB constraints to validate any hints of dynamical dark energy.

Abstract

We investigate the impact of instrumental and astrophysical systematics on dark energy constraints derived from Type~Ia supernova (SN-Ia) observations. Using simulated datasets consistent with current SN-Ia measurements, we explore how uncertainties in photometric calibration, intergalactic dust, progenitor evolution in luminosity and light-curve stretch, and intrinsic color scatter affect the inferred dark energy equation of state parameters (w0, wa). We test the Generalised Scale Factor (GEN) evolution and benchmark it against three time-evolving dark energy models; namely Chevallier Polarski Linder (CPL), Jassal Bagla Padmanabhan (JBP) and Logarithmic (LOG) parametrizations; comparing their sensitivity to these systematic effects. Calibration biases and progenitor evolution emerge as the dominant sources of uncertainty, while simpler parametrisations, viz. GEN, which directly describes the expansion rate, remains relatively stable under all systematic injections, unlike CPL, JBP and LOG that rely on the dark energy equation of state. These findings underscore the need for sub-per cent calibration precision and enhanced astrophysical modelling to ensure the robustness of dark energy inferences from current and future SN-Ia cosmology experiments.

Figures (5)

• Figure 1: Impact of injected SNIa systematic uncertainties on the Hubble residuals, illustrating how each source of systematics alters the inferred cosmological distances.
• Figure 2: Impact of injected Type Ia supernova systematics on cosmological parameter constraints from the CPL vs JBP models. Each panel shows the shift in the $w_0$--$w_a$ plane resulting from a specific injected systematic (e.g., calibration, color correction, progenitor effects (on $M$ and $x_1$), intergalactic dust, and $\Omega_m$ mismatch). Filled contours correspond to fits without systematics, while dashed contours indicate results after applying the injected systematic, scaled for better visualisation.
• Figure 3: Impact of injected SNIa systematics on the $w_0$--$w_a$ constraints for the CPL vs LOG parametrizations. Panels illustrate how individual systematics shift the recovered contours. Filled contours denote fits without systematics; dashed contours show results after systematic injection.
• Figure 4: Impact of injected SNIa systematics on the $w_0$--$w_a$ constraints for the CPL vs GEN parametrizations. Panels illustrate how individual systematics shift the recovered contours. Filled contours denote fits without systematics; dashed contours show results after systematic injection.
• Figure 5: Effect of systematics injection on the normalised Hubble parameter $E(z)$ (left), distance modulus $\mu(z)$ (middle) and dark energy EoS $w(z)$ (right): Comparison between CPL (top panel) vs GEN (bottom panel) models.