A web platform from the NASA International Space Apps Challenge 2025 for monitoring and forecasting air quality using real-time environmental data and predictive models.
- Use TEMPO (Tropospheric Emissions: Monitoring of Pollution) satellite data combined with ground-based air quality measurements and weather data to forecast air quality.
- Alert users to bad air conditions (high AQI).
- Support communities and health agencies in making health-protective decisions.
- Use cloud computing to scale from local devices to global cloud systems.
- Real-time data ingestion: automatically get data from TEMPO, NOAA weather, ground stations.
- Machine Learning forecasting: use spatio-temporal models (LSTM, Temporal Convolutional Networks, Ensemble models) to predict AQI.
- Uncertainty quantification: provide confidence intervals for forecasts.
- Web dashboard: interactive map showing air quality over time & space.
- Notification system: send alerts when AQI exceeds WHO/EPA thresholds.
- Cloud-native scaling: run on Kubernetes/Serverless functions to accommodate thousands of users.
- TEMPO satellite data
- EPA AirNow ground station data
- NOAA weather forecasts
- WHO Ambient Air quality database
- Baseline: ARIMA, Linear Regression.
- Deep Learning: LSTM, TCN, Transformer-based time-series models.
- Ensemble:
- Evaluation Metrics: RMSE, MAE, R-squared