April 2026 CACM: Where Are the City Trees? Monitoring Urban Trees Across the U.S. Using GenerativeAI

Why It Matters

Accurate, city‑wide tree inventories empower municipalities to combat heat islands, lower energy use, and enhance public health through data‑driven urban forestry decisions.

Key Takeaways

• AI-driven satellite mapping creates nationwide tree inventories quickly
• Method achieves 92% accuracy in tree count, 1.5 m location precision
• Three-phase workflow: monthly imagery, category classification, generative AI refinement
• Scales to 330 cities, reducing analysis time from months to days
• Enables policymakers to target heat islands and improve air quality

Summary

Urban planners and researchers unveiled a generative‑AI system that leverages monthly satellite imagery to inventory every tree across U.S. cities. The approach replaces labor‑intensive field surveys with a three‑phase workflow: (1) capture twelve monthly images to track phenological changes, (2) classify canopies into roadside, residential, industrial/commercial, and park categories, and (3) apply generative AI to resolve individual trees obscured in the imagery.

Validation against 19 detailed and 80 semi‑detailed municipal inventories showed 92 % accuracy in tree counts and a mean positional error of 1.5 m, scaling the analysis to roughly 330 cities in a fraction of the time traditional methods require.

The researchers aim to launch a national tree‑inventory portal, giving cities actionable data to target heat‑island mitigation, improve air quality, and reduce energy consumption.

By democratizing high‑resolution urban forest data, the tool promises smarter, greener city planning and more resilient communities.

Original Description

Daniel Aliaga and Adnan Firoze discuss "Where Are the City Trees? Monitoring Urban Trees across the U.S. Using Generative AI," a Research Article in the April 2026 CACM.