In collaboration with Don Lamb, Rick Kessler, and the astrophysics group at the FLASH center, I evaluate simulations of Type Ia Supernovae by comparing their light curves with observations. Below I describe some of our papers, in reverse chronological order.
Three-dimensional simulations of pure deflagration models for thermonuclear Supernovae
We present six very high resolution hydrodynamical simulations of so-called pure deflagration models for Type Ia Supernovae (SNeIa). In such models, only a deflagration (sub-sonic nuclear flame) occurs, but no detonation (super-sonic flame). Thus, these models produce relatively small explosions and cannot explain most of the SNeIa we observe. The figure below shows the initial white dwarf (blue), and the nuclear flame at times between zero seconds (left) and 2 seconds.
Since the initial distribution of burning bubbles (orange dots on the left) is unknown, we investigate its influence on the outcome of the explosion. We find that in different stages of the explosion the nuclear burning rate depends on the number of bubbles, their density, and the size of the sphere they are arranged in. We carefully compare the light curves predicted by our simulation results to observations. We find that some of our pure deflagration models may account for under-luminous events.
Comparing the light curves of simulated Type Ia Supernovae
with observations using data-driven models
Since SNe Ia are quite heterogeneous in their light curves and spectral properties, there is no simple template to compare simulations to. Our strategy is to use the data-driven model SALT2 to make inferences about how well the simulated light curves match observed data. The SALT2 model summarizes information from a large set of observed SNe Ia and parametrizes them with three free parameters: stretch, color, and magnitude.
We have derived a statistical description of the underlying population of observed SNe Ia in this stretch-color-magnitude parameter space. We then fit explosion models with SALT2, and compare the fitted parameters to our population model. We use two metrics to quantify how well a model agrees with observations: the quality of the light curve fit, and how likely it is to observe a SNe Ia with the fitted stretch, color and magnitude in nature.
The figure shows a fit of the SALT2 model (dashed blue lines) to the W7 explosion model (Nomoto et al. 1984, black lines) in four filter bands. The fit results (stretch, color and magnitude) are shown in the top panel.