Evolve AI Institute

Case Study 2: IBM Green Horizon for Air Quality Prediction

Lesson 6: AI in Climate Science and Prediction

IBM's Green Horizon for Air Quality Prediction

The Challenge

Air pollution kills 7 million people worldwide annually according to the World Health Organization. Cities struggle to warn citizens about dangerous air quality in time for people to take protective measures. Traditional air quality monitoring relies on limited ground sensors and can't predict pollution levels accurately more than a few hours in advance.

The AI Solution: IBM Green Horizon

Technology: AI-powered air quality forecasting system

Developed by: IBM Research with partners in China, India, and other countries

How it works:

Prediction Accuracy:

Real-World Impact

Public Health Protection:

City Management:

Economic Benefits:

Notable Success:

The Data Behind the AI

Input Sources:

AI Training:

Machine Learning Approach:

Critical Thinking Questions

For Your Group Discussion:

  1. Complexity of Predictions: Air quality depends on dozens of variables. How does AI handle this complexity better than traditional methods? What makes 72-hour prediction so challenging?
  1. Data Privacy: The system tracks traffic patterns and industrial activity. What privacy or business confidentiality concerns might arise? How can these be balanced with public health needs?
  1. Action Threshold: At what pollution level should cities take action? Who decides? What if the AI prediction is wrong and cities unnecessarily restrict traffic or industry?
  1. Environmental Justice: Poor neighborhoods often have worse air quality. How can AI predictions help address these inequities? Could the technology make inequities worse?
  1. Verification: How do scientists verify that AI predictions are accurate? What happens when actual pollution differs from predictions?
  1. Behavioral Change: If people receive daily air quality predictions, will it change their behavior? What psychological factors influence whether people act on warnings?
  1. Root Causes: This AI predicts pollution but doesn't eliminate sources. How should cities use AI insights to reduce pollution long-term, not just manage short-term spikes?
  1. Climate Connection: How is air pollution related to climate change? Can addressing one help address the other?

Your Analysis Task

Complete the following for your case study presentation:

Summary (2-3 sentences):

Key Data & Methods:

Environmental Problem Addressed:

Real-World Impact:

Limitations or Concerns:

Climate Change Connection:

One Question for Class Discussion:

Real-World Scenario

Beijing, China - January 2024

The IBM Green Horizon system predicted severe air pollution would develop on Thursday due to:

City Response Based on AI Prediction:

Result:

Discussion: Was this the right decision? What if the prediction had been wrong?

Additional Resources

Learn More:

Related AI Applications:

Key Pollutants Monitored:

Vocabulary

PM2.5: Particulate matter smaller than 2.5 micrometers (1/30th width of human hair) - penetrates deep into lungs

Temperature Inversion: Weather condition where warm air traps cool air and pollutants near ground

Ensemble Modeling: Using multiple AI models together to make more accurate predictions than any single model

Aerosol Optical Depth: Measure of how much sunlight pollution particles block - visible from satellites

Dispersion Model: Computer simulation of how pollutants spread through atmosphere based on wind and weather

© 2025 Evolve AI Institute LLC. May be reproduced for educational purposes.

Lesson 6: AI in Climate Science | Case Study 2 of 6