Fires generate and transport embers in complex turbulent air columns. This air mixture contains wide ranges of air velocities, transporting embers farther than classical plume models. Frequently, wildfire models make assumptions like steady-state plumes and continuously released ember generation, potentially ignoring the chaotic and intermittent fire physics. This study uses high-speed imaging to track ember production and trajectories from a pile burn experiment, calculating a distribution of ember generation rate and velocity. The experimental data is then incorporated into a simulation to analyze how accounting for these random processes affect ember transport and fallout. We discuss how this comparison can be parameterized and implemented into wildfire models to better predict how a fire will spread.