Measuring the extent and vertical structure of urban tree canopies has long relied on ground-level surveys — slow, expensive, and difficult to replicate at metropolitan scale. Over the past five years, the combination of freely available satellite imagery, airborne LiDAR datasets, and machine learning inference has shifted that calculus substantially in Italy's two largest cities.
Milan: Forestami and the 10-Million-Tree Inventory
The Forestami project, initiated in 2019 through a formal agreement between Politecnico di Milano, the Municipality of Milan, and the Metropolitan City of Milan, produced the first systematic canopy inventory for the metropolitan territory. The mapping methodology cross-referenced municipal land-use layers with satellite data and field sample plots, arriving at an estimated 10 million trees within the metropolitan boundary.
That figure — covering a territory of approximately 1,575 km² — placed canopy cover at 16% of total land area as measured in the 2018–2020 baseline period. The project operates across three phases: initial analysis concluded in 2020; metropolitan-scale development mapping ran from 2021 to 2023; and the current phase through 2026 focuses on monitoring planted cohorts and refining the GIS platform.
Sentinel-2 and Deep Learning Canopy Height Models
A 2024 study published through the Transform Transport research network evaluated two global canopy height models against ground-truth species location data from the Comune di Milano tree registry. The ETH Global Sentinel-2 10m Canopy Height dataset (2020) and Meta's Global Canopy Height Maps were compared using Sentinel-2 imagery captured on 19 June 2024 and NDVI-derived green coverage calculations.
The findings highlighted systematic underestimation of canopy height in dense street corridors, where building shadows and mixed pixels obscure tree crowns from optical sensors. LiDAR integration was identified as necessary to resolve height ambiguity below 8 metres — the typical crown height of recently planted street trees in the city's planting cohorts.
Rome: Street Tree Master Plan and Spatial Data
Rome published a formal Street Tree Master Plan in 2024, the first document of its kind to incorporate GIS-derived canopy data as a baseline for species-rotation scheduling. The plan, developed by the Dipartimento Tutela Ambientale, draws on the city's existing tree registry — which records approximately 300,000 managed trees — and supplements it with remote sensing analysis of unmapped roadside vegetation.
The Latium regional forest mapping study published in early 2025 found that trees (including urban fringe stands) cover 44.2% of the total regional land area. The study, conducted using the CORINE Land Cover classification updated with regional satellite composites, distinguished between closed-canopy stands, linear plantings, and scattered individual trees — a distinction that prior regional inventories had not formalised.
The GIS Workflow in Municipal Practice
Both cities rely on a layered approach. Base layers come from the national topographic map (CTR) and cadastral boundaries. Tree point data — derived from either field survey or satellite-based detection — are joined to species records and health assessments. Change detection between successive satellite composites then flags areas of canopy loss or gain for field verification.
Rome's Greenspaces monitoring platform, referenced in the city's 2024–25 planting season announcements, integrates real-time maintenance records with the spatial tree inventory. Pruning interventions (26,000 scheduled for the first quarter of 2025 alone) are logged against individual tree IDs, enabling longitudinal health tracking at scale.
Limitations and Open Data
Neither Milan nor Rome has published a fully open canopy coverage dataset at the resolution produced by their internal monitoring systems. The Forestami project makes summary statistics available on its public-facing website but does not release the underlying shapefile or raster products. Rome's tree registry is accessible via the open data portal of Roma Capitale but does not include canopy polygon geometries — only point locations with associated attribute tables.
The gap between what municipal GIS systems hold and what is released for external research and verification remains one of the structural constraints on independent canopy monitoring in Italian cities. Several researchers involved in the cited studies have called for mandatory publication of canopy datasets as a condition of EU-funded urban greening grants.