Spurious Ionization From Noise in Kinetic Particle Simulations

The inclusion of field ionization models in kinetic plasma simulations is desired in several applications. The ionization rates in these models are sensitive to noise in the electric field; however, and in some cases, spurious "numerical" ionization may dominate the physical ionization. In this paper, analytic ionization rates are obtained as a function of the noise spectrum of the electric field. These expressions have applicability to several settings, including colored noise, as are obtained from digital filters. As such, they provide a means for quantitatively assessing the efficacy of candidate noise mitigation schemes, such as increasing the computational particle number, temporally averaging the electric field, or applying more sophisticated digital filtering techniques