Data Collection:
- Historical Data: Gather data on past meteor showers, fireball events, and the parameters that affected their visibility and intensity (e.g., date, time, geographic location, atmospheric conditions).
- Astronomical Data: Include information about the orbits of meteoroids, space objects, and celestial phenomena that might influence the likelihood of meteor showers.
- Real-time Data: Integrate weather patterns, atmospheric transparency, and space environmental factors to improve predictions.
- Data Preprocessing: Clean and preprocess historical and real-time data, including outlier removal and feature scaling.
- Model Selection: Implement machine learning models such as time series forecasting algorithms (e.g., ARIMA, LSTM) and classification models to predict both the time and location of meteor events.
- Model Training: Train the AI models on historical meteorological and astronomical data to understand the patterns that indicate upcoming meteor showers or fireballs.
- Model Validation: Validate predictions using real-time data to ensure accuracy and reliability.
- Create a system to provide visualizations and alerts when certain thresholds of probability for meteor showers or fireball occurrences are met.
- Implement a user-friendly interface that can communicate predictions to researchers, enthusiasts, or the public.
- Partner with astronomical observatories and amateur skywatching communities for validation and real-time event reporting.
- Use crowdsourced data (e.g., from meteor observation apps) to improve model performance.

2024 NASA Space Apps Challenge

ARTAN