Radial Diffusion Benchmarking: Initial Conditions

Earth’s radiation belts are a hazardous environment containing trapped charged particles. Radial diffusion is one of the major processes driving radiation belt physics, accounting for energisation, transport and loss of electrons in the outer belt. The outer radiation belt is highly variable in energy and location, resulting in behaviour which is difficult to model accurately.

 

Ensemble modelling is needed to characterise this variability. Ensembles can be constructed by varying physical parameters (capturing the uncertainty in our knowledge across many scales) and considering the spread of the final model outputs. However, it is unclear what proportion of the subsequent variability comes from physics versus the numerical methods used. We investigate the effect of varying initial conditions for typical radial diffusion coefficients.

 

We present two methods of establishing the timescale over which initial conditions affect the subsequent radial diffusion; time to monotonicity (the time taken for the particle distribution to reach a state where radial diffusion effects become uninteresting) and dimensional analysis. Both are needed to capture the processes we are interested in as well as the inherent timescales from diffusion. Our measures are often domain dependent, indicating that the choice of where we perform our radial diffusion simulations is significant.