Modeling collision probability for Earth-impactor 2008 TC3

TL;DR

This paper models the collision probability of asteroid 2008 TC3 with Earth using probabilistic orbital methods and observational data, providing insights into impact risk assessment with limited data.

Contribution

It introduces a methodology combining Markov-chain Monte Carlo and Monte Carlo ranging to evaluate impact probability from short observational arcs.

Findings

Orbital uncertainties decrease over time after discovery.

Impact probability is sensitive to assumed astrometric noise.

Both MCMC and Monte Carlo methods yield comparable impact probabilities.

Abstract

We study the evolution of the Earth collision probability of asteroid 2008 TC3 using a short observational arc and small numbers of observations. To assess impact probability, we use techniques that rely on the orbital-element probability density function characterized using both Markov-chain Monte-Carlo orbital ranging and Monte-Carlo ranging. First, we evaluate the orbital uncertainties for the object from the night of discovery onwards and examine the collapse of the orbital-element distributions in time. Second, we examine the sensitivity of the results to the assumed astrometric noise. Each of the orbits obtained from the MCMC ranging method is propagated into the future (within chosen time bounds of the expected impact), and the collision probability is calculated as a weighted fraction of the orbits leading to a collision from the Earth. We compare the results obtained with both methods.