Light curve spectral energy distribution fitting of 13 carbon-rich Mira variables stars in the Small Magellanic Cloud

Abstract

Carbon-rich Mira variables are highly dynamic stars characterized by large-amplitude radial pulsations and intense mass-loss. Standard spectral energy distribution (SED) fitting often relies on time-averaged observations, which flatten intrinsic variability and yield only approximate stellar and circumstellar parameters. To overcome these limitations, we dynamically trace the physical parameters of 13 carbon-rich Miras in the Small Magellanic Cloud (SMC) throughout their pulsation cycles. Combining time-series photometry from $Gaia$, WISE, VMC, 2MASS, and OGLE, we reconstruct simultaneous multi-band light curves. We extract independent SEDs at 14 distinct phases and model them using the ATHENA-C grid. This phase-resolved approach traces cyclic variations in effective temperature ($T_{\rm eff}$), optical depth ($τ_{0.55}$), and bolometric luminosity. We find peak-to-peak $T_{\rm eff}$ variations ranging from $\sim 100$ K to $>1000$ K. A strong anti-correlation between $τ_{0.55}$ and bolometric luminosity indicates the dust condensation zone is pushed outward at maximum brightness, rendering the circumstellar environment optically thinner. We also recover the classical Period-Luminosity relation and derive radial expansion velocities and radius variation amplitudes, which are consistent with independent measurements. Phase-resolved fitting reveals the true dynamic ranges of these parameters, which static fits fail to capture. Although limited by grid sampling and gaps where converged hydrostatic models are unavailable, explicitly accounting for time-dependent behavior provides a much more physically realistic picture of the interplay between stellar pulsation and dust formation. This highlights the critical need for time-resolved methodologies when studying large-amplitude variables.