Beyond the curb — Toronto's actual canopy, and where the equity gap really lives

In an earlier post I looked at canopy equity using the city's street-tree inventory — the trees on the boulevard between your sidewalk and the road. There was a strong correlation between income and street-trees-per-capita (r = 0.67), and a 28% gap between Neighbourhood Improvement Areas and the rest of the city.

A fair critique of that analysis: the inventory only covers trees the city owns on road allowances. It leaves out parks, ravines, institutional grounds, and — the big one — private property. In most Toronto neighbourhoods that's where the majority of actual tree cover lives.

Toronto publishes a 2018 land-cover study, derived from LiDAR and aerial imagery, that classifies every square metre of the city into one of eight categories: tree canopy, shrub, grass, bare earth, water, buildings, roads, other paved. Summed over a neighbourhood, it gives you the actual percent of the ground shaded by tree crowns from above — not what the city planted, but what's growing.

Here it is:

Citywide canopy cover in 2018: 25.9%. The City of Toronto's Strategic Forest Management Plan targets 40% by 2050. We're two-thirds of the way there.

Top and bottom

A few notable patterns:

• Ravine adjacency dominates the top. Bridle Path (the Don Valley), Rosedale-Moore Park (Rosedale Ravine), Lambton Baby Point (Humber River), High Park-Swansea (High Park itself). If there's a ravine next door, you have canopy.
• High-rise cores dominate the bottom. Yonge-Bay Corridor, Harbourfront-CityPlace, Wellington Place — the new condo core is essentially treeless from above.
• Cabbagetown at 45% is a genuine surprise. Victorian-era single-family streets, backyard mature canopies, and a sliver of Riverdale Park adjacent. It's a middle-income canopy standout, $76k median vs. Rosedale's $122k.
• Downsview at 8% is the lowest NIA. The former airbase is still mostly grassland and runway paths — its "green" colour on maps is grass, not trees.

The equity re-read

Here's the key comparison. On the left: canopy cover vs. income. On the right: street trees per capita vs. income. Same neighbourhoods, same income data, different tree metrics.

Both relationships are real. But they're not the same strength.

• Total canopy vs. income: r = 0.50
• Street trees per capita vs. income: r = 0.67

The street-tree relationship is tighter than the total-canopy relationship. Which is the opposite of what you might expect — you'd assume the "bigger" measure (all trees everywhere) would show a stronger income relationship, because wealthy homeowners with big backyards plant trees.

They do. But the city plants even more trees in front of their houses.

The street-tree gap is bigger than the total-canopy gap

Indexed to "Not NIA or Emerging" neighbourhoods = 100:

NIAs have 11% less total canopy than non-NIAs. But they have 28% fewer street trees per capita. The public-land gap is about 2.5× larger than the private-land gap.

This is a genuinely different equity framing than "tree equity" as it's usually discussed. A standard canopy analysis — the kind i-Tree studies and most municipal forestry reports do — would conclude Toronto has modest canopy inequity (which, at r = 0.50, it does). The policy-relevant framing is narrower: on the land the city directly controls, it plants markedly fewer trees per resident in low-income neighbourhoods. That's a tractable problem.

One more way to see it: for every 1,000 m² of actual tree canopy in a neighbourhood, how many street trees does the city maintain?

Emerging neighbourhoods are actually above the non-NIA average on this — they're mostly newer suburbs where the city's street-tree program has rolled out but the private canopy hasn't matured yet. NIAs are 12% below average: their canopy is more dependent on non-city sources (yards, parks, ravines) than the wealthy-neighbourhood canopy is. If anything were to happen to those non-city sources — infill development, institutional land changes — the damage would land hardest where the city's street-tree cushion is thinnest.

Caveats

• The 2018 land-cover data classifies "tree canopy" as woody-plant tree-form cover. Shrub is a separate class. Including shrub would bump city-wide cover from 25.9% to about 27.8%. I used tree-only to match how municipal reports usually state canopy.
• Neighbourhood boundaries are the 158-model (current). The 2018 data is already dissolved by neighbourhood in the geodatabase — no spatial re-joining was needed on my end.
• Correlation isn't causation. High-income neighbourhoods often have older tree stock, more backyard space, and ravine proximity, all of which compound. Separating the contributions is another analysis.

What comes next

Toronto also publishes a 2008 land-cover study using the same classification scheme. Diffing 2008 → 2018 would show which neighbourhoods gained canopy and which lost it over a decade that included the emerald ash borer devastation. I'll write that one next.