Velocity dispersion of Solar Energetic Particles in turbulent heliosphere

TL;DR

This study uses simulations to evaluate how turbulence affects velocity dispersion analysis (VDA) in determining solar energetic particle injection times, revealing significant influences from turbulence and background spectra.

Contribution

The paper introduces a comprehensive simulation framework to assess the impact of magnetic turbulence on VDA estimates of SEP injection times and path lengths.

Findings

Turbulence causes VDA to overestimate injection times by 2-16 minutes.

Path lengths derived from VDA are significantly longer in strong turbulence, exceeding 5 au.

Background spectra variations lead to 5-20 minute differences in VDA injection times.

Abstract

Solar Energetic Particles (SEPs) are a signature of solar eruptions, and to link them to acceleration mechanisms many studies investigate their injection time at the Sun, $t_{sun}$. We assess velocity dispersion analysis (VDA), an often-used method to derive $t_{sun}$. We use full-orbit simulations of 1--100 MeV SEP protons in a novel model of the interplanetary magnetic turbulence superposed on a Parker Spiral magnetic field. The turbulence is described analytically as dominant transverse fluctuations that are 2D with respect to the mean field, supplemented with a minor contribution of asymptotically slab turbulence modes. We determine simulated SEP intensities for three turbulence strengths and use VDA to obtain $t_{sun}$ and the apparent path length $s$ of the SEPs, employing an SEP onset threshold to mimic a realistic energetic proton background before the SEP event. We find that turbulence strongly affects $t_{sun}$ and $s$. For weak and moderate turbulence, VDA estimates of $t_{sun}$ are 2-16 minutes after the actual solar injection time, and the path lengths are 0.2-0.3 au longer than the Parker spiral. For strong turbulence, the path lengths are $>5$ au, considerably longer than those typically obtained from SEP observations. We also investigate the effect of energy-dependence of the pre-event proton background, and find that different background spectra result in 5-20-minute difference in VDA injection times, depending on the heliolongitude. We conclude that in many cases VDA-derived injection times include a significant contribution from turbulence and/or the pre-event background and are not an accurate estimate of the acceleration time.