Warning Times and Impact Predictions of Asteroids and Comets on a Collision Course with Earth

This study investigates the amount of data and time necessary to accurately predict Earth impacts of Earth Crossing Objects (ECOs). Trajectories are simulated by numerically integrating in an N-Body system. Given final impact parameters, the trajectory is propagated backwards to an earlier time, creating initial conditions and simulated observation data at requested intervals to which Gaussian random noise is introduced. Utilizing a Bayes Filter to estimate position and velocity from the simulated observation data, the estimate is then propagated forward in time to determine whether or not an impact can be accurately predicted. State vectors and covariance matrices are then propagated to the impact time and the one sigma error ellipsoid is analyzed