# Biennial sub-meter tree coverage dataset of Orlando (2013–2021)

## TL;DR

This paper presents a high-resolution dataset tracking tree coverage in Orlando from 2013 to 2021 to support urban planning and environmental sustainability.

## Contribution

The novel contribution is a validated, biennial sub-meter tree coverage dataset for Orlando with improved accuracy over existing products.

## Key findings

- Tree detection accuracy ranged from 78% to 89% through extensive ground-truth validation.
- The dataset addresses sensor inconsistencies and environmental challenges in urban tree mapping.
- The dataset supports urban forest management and broader applications in planning and public health.

## Abstract

Rapid urban development often comes at the cost of significant vegetation loss, and the loss of urban trees is a particularly concerning issue. Many cities have recognized the crucial role of tree canopy cover in addressing a range of pressing environmental challenges, and have set goals in their development plans. However, the lack of resources for monitoring its spatiotemporal changes directly undermines efforts of cities to balance urban growth with environmental sustainability. We developed high-resolution biennial tree canopy maps (2013–2021) for the Orlando metropolitan area using National Agricultural Imagery Program (NAIP) orthoimagery. Our approach addresses challenges such as sensor inconsistencies, shadows, variable observation times, and environmental factors. Extensive ground-truth validation demonstrates 78%–89% accuracy in tree detection, outperforming existing products. This dataset provides the crucial link between planning and action, and supports urban forest management and has broader applications in urban planning, environmental and public health monitoring, and education programs.