Abstract:Risk assessment studies of local asteroid hazards traditionally simulate the physics of meteors with engineering models tailored to analyze tens-of-millions of scenarios. However, these simplified approaches still need to solve time-dependent ODEs to model the entry process and the resulting ground damage. With a computational cost of O(0.01 CPU.s) per scenario, simulating these large numbers of potential entry conditions can take several days on local computers. To improve computational efficiency, we propose in this paper an orthogonal approach based on machine learning models to predict the size of damaged areas given a list of entry parameters. We train 5 machine learning methods and compare the predictions to the outputs of NASA's state-of-the-art Probabilistic Asteroid Impact Risk (PAIR) model. PAIR's remarkable capability to handle a broad spectrum of entry conditions and a high volume of scenarios establishes it as our reference model, with estimates assumed ground truth and used to train, test, and validate the machine learning models. The ML models, initially trained with PAIR's primitive input variables, are then further enhanced with engineered features to improve the performance of interpretable models. We find that complex models like neural networks are well-suited to estimate blast hazards, while simpler, interpretable linear models can accurately assess thermal damage. For both types of hazards, the radii of damaged areas can be predicted with less than 10% average relative errors for radii greater than 100km. The CPU time is decreased by a factor O( 103 ) compared to the PAIR model, enabling the simulation of millions of scenarios within minutes on a local computer. Beyond serving as surrogate models, we incorporate machine learning algorithms into a comprehensive Shapley sensitivity analysis, yielding a ranking of entry parameters based on their contributions to ground damages. These results offer valuable insights for prioritizing observation missions.

Prediction of Apophis Asteroid Flyby Optimal Trajectories and Data Fusion of Earth-Apophis Mission Launch Windows using Deep Neural Networks

Eclipse Prediction and Orbit Improvement for Asteroids: Theory and Application to Near Earth Asteroids

Initial Trajectory Assessment of a low-thrust option for the RAMSES Mission to (99942) Apophis

Initial Trajectory Assessment of the RAMSES Mission to (99942) Apophis

Fast Solution Continuation of Time-Optimal Asteroid Landing Trajectories Using Deep Neural Networks

Influence of Apophis' spin axis variations on a spacecraft during the 2029 close approach with Earth

Fast Generation of Optimal Asteroid Landing Trajectories Using Deep Neural Networks

Temporal trends in asteroid behaviour: a machine learning and N -body integration approach

Advancement and Challenges in Methods of Predicting the Motion of Near-Earth Asteroids

DNN Estimation of Low-Thrust Transfer Time: Focusing on Fast Transfers in Multi-Asteroid Rendezvous Missions

Machine learning for the prediction of local asteroid damages

Optimal Combined Impulsive/Low-Thrust Trajectories for Asteroid Deflection Via Kinetic Impact

A deep neural network framework with Analytic Continuation for predicting hypervelocity fragment flyout from satellite explosions

How selection and weighting of astrometric observations influence the impact probability. Asteroid (99942) Apophis case

Real-time optimal control for irregular asteroid landings using deep neural networks

A novel approach to asteroid impact monitoring and hazard assessment

Solar-Sail Trajectory Design of Multiple Near Earth Asteroids Exploration Based on Deep Neural Network.

First approximation for spacecraft motion relative to (99942) Apophis

On the sensitivity of Apophis' 2029 Earth approach to small asteroid impacts

Autonomous Rapid Exploration in Close-Proximity of an Asteroid

Revisiting Apophis 2029 approach to Earth (staying on shoulders of NASA experts) or Can we be sure in almost ricocheting fly-by of Apophis on 13 of April 2029 near the Earth?