Bayesian inference of β-meteoroid parameters with Solar Orbiter

Solar Orbiter’s Radio and Plasma Waves instrument (SolO/RPW) is capable of detecting hypervelocity dust impacts onto the spacecraft through the fast electrical phenomena that accompany the process. SolO operates within 1AU, in the environment with high density of β-meteoroids – dust grains escaping from the proximity of the Sun due to radiation pressure force counteracting gravity. Recently, Convolutional Neural Network (CNN) classified data were made available^[1], analyzing all the recorded waveforms and providing us with the highest quality dataset of the impact events to date.

We present a model for the in-situ impact rate on SolO/RPW assuming β-meteoroids are the main component of the detections. We fit the model to the highest quality available CNN data assisted by Integrated Nested Laplace Approximation (INLA) for Bayesian inference with informative priors^[2].

Taking into account spacecraft’s position and its velocity vector, we are able to infer mean radial velocity of the detected dust grains to be 63 ± 7 km/s. We are also able to constrain β-meteoroid predominance and dust’s mean acceleration and by extension constrain its mean β-parameter. The procedure is general enough to be used in a different setting for SolO, or by a different spacecraft in the future.

References:

[1] Kvammen, Andreas, et al. "Machine Learning Detection of Dust Impact Signals Observed by The Solar Orbiter." (2022). https://doi.org/10.5194/egusphere-2022-725

[2] Kočiščák, Samuel, et al. "Modelling Solar Orbiter Dust Detection Rates in Inner Heliosphere as a Poisson Process." arXiv preprint arXiv:2210.03562 (2022). https://doi.org/10.48550/arXiv.2210.03562