NASA Zeus Air Quality Monitoring System

Real-time Air Quality Intelligence with AI-Powered Predictions

Data Sources

Source	Type	Usage
OpenAQ	Ground Stations	Real-time PM2.5, PM10, O3, NO2, SO2, CO
NASA TEMPO	Satellite	Tropospheric air quality observations
NASA MERRA-2	Reanalysis	Historical atmospheric data (15,552 records)
NOAA GFS	Model	Weather forecasts and atmospheric parameters
NOAA METAR	Ground Stations	Surface temperature and pressure

---

Overview

NASA Zeus is an enterprise-grade air quality monitoring platform that combines real-time ground station data, satellite observations, and machine learning to provide actionable air quality intelligence. Developed for the NASA Space Apps Challenge, it addresses critical environmental monitoring needs through innovative data integration and AI-powered analytics.

Key Capabilities

• Real-Time Monitoring: Interactive heat maps with live air quality data from 10,000+ stations worldwide
• Satellite Integration: NASA TEMPO and MERRA-2 satellite data for comprehensive atmospheric analysis
• AI Weather Agent: Google Gemini-powered intelligent assistant for atmospheric data retrieval and analysis
• Predictive Analytics: XGBoost-based machine learning model for O3 (ozone) concentration predictions
• Multi-Source Fusion: Seamless integration of ground stations, satellite data, and weather forecasts
• Enterprise Security: JWT-based authentication with role-based access control
• Historical Analysis: Time-series visualization and trend analysis for pollution patterns
• Real-Time Alerts: Configurable threshold-based notifications for air quality events

---

Architecture

┌─────────────────────────────────────────────────────────────────┐
│                        Client Layer                              │
│  ┌──────────────────────────────────────────────────────────┐   │
│  │         Next.js Frontend (React + TailwindCSS)            │   │
│  │  • Interactive Maps (Leaflet) • Real-time Dashboard       │   │
│  │  • User Authentication • AI Chat Interface                │   │
│  └──────────────────────────────────────────────────────────┘   │
└─────────────────────────────────────────────────────────────────┘
                              ▼
┌─────────────────────────────────────────────────────────────────┐
│                     Application Layer                            │
│  ┌─────────────────────┐  ┌──────────────────────────────────┐  │
│  │   FastAPI Backend   │  │   Gemini AI Service              │  │
│  │   • REST API        │  │   • Atmospheric Data Retrieval   │  │
│  │   • Authentication  │  │   • O3 Prediction Engine         │  │
│  │   • Data Aggregation│  │   • XGBoost ML Model             │  │
│  │   Port: 8000        │  │   Port: 8001                     │  │
│  └─────────────────────┘  └──────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────┘
                              ▼
┌─────────────────────────────────────────────────────────────────┐
│                      Data Layer                                  │
│  ┌──────────┐  ┌──────────┐  ┌──────────┐  ┌──────────┐        │
│  │ OpenAQ   │  │   NASA   │  │  NOAA    │  │ Weather  │        │
│  │   API    │  │  TEMPO   │  │   GFS    │  │   APIs   │        │
│  └──────────┘  └──────────┘  └──────────┘  └──────────┘        │
│       Ground Stations    Satellite Data    Atmospheric Models   │
└─────────────────────────────────────────────────────────────────┘

---

Features

User Interface

• Responsive Design: Mobile-first approach with TailwindCSS
• Interactive Maps: Leaflet-based heat maps with real-time data overlay
• Dark/Light Mode: Customizable theme preferences
• Accessibility: WCAG 2.1 AA compliant

Data & Analytics

• Multi-Pollutant Tracking: PM2.5, PM10, O3, NO2, SO2, CO monitoring
• Historical Data: Access to years of archived measurements
• Trend Analysis: Statistical analysis and visualization
• Data Export: CSV, JSON formats for research use

AI & Machine Learning

• Gemini Integration: Natural language queries for atmospheric data
• O3 Prediction Model:
  • XGBoost regression model
  • Input features: PS, TS, CLDPRS, Q250, TO3
  • Trained on 15,552 NYC MERRA-2 records
  • Prediction accuracy: RMSE < 15 ppb

Security & Authentication

• JWT Tokens: Secure stateless authentication
• Password Hashing: bcrypt with salt rounds
• CORS Protection: Configurable cross-origin policies
• Rate Limiting: API throttling to prevent abuse

---

API Documentation

Authentication Endpoints

Method	Endpoint	Description
POST	/auth/register	Create new user account
POST	/auth/login	Login and receive JWT token
GET	/auth/me	Get current user info

Air Quality Endpoints

Method	Endpoint	Description
GET	/api/air-quality	Get real-time air quality data
GET	/api/stations	List all monitoring stations
GET	/api/historical/{location}	Get historical data

Gemini AI Endpoints

Method	Endpoint	Description
GET	/atmospheric-data?location={city}	Get atmospheric parameters
GET	/predict-o3?location={city}	Predict O3 concentration

---

Technology Stack

Frontend

• Framework: Next.js 15 (React 19)
• Styling: TailwindCSS
• Maps: Leaflet + React-Leaflet
• State Management: React Hooks
• HTTP Client: Axios

Backend

• Framework: FastAPI (Python 3.9+)
• Database: SQLAlchemy + SQLite
• Authentication: JWT (python-jose)
• Password Hashing: bcrypt
• Validation: Pydantic

AI & Machine Learning

• LLM: Google Gemini 1.5 Flash
• ML Framework: XGBoost 1.7.6
• Data Processing: Pandas, NumPy
• Scientific: SciPy, scikit-learn

DevOps

• Cloud: AWS EC2 (t3.small)
• CI/CD: GitHub Actions (optional)
• Monitoring: CloudWatch (optional)
• Containerization: Docker (optional)

---