Jacobs vs Wilderness
Two ways of measuring a park, deliberately held in tension. Urban integration (the x-axis) reads parks through Jane Jacobs’ lens. The model averages edge activation, connectivity, and enclosure to ask does this park live inside the city? Natural comfort(the y-axis) reads them through a forester’s lens, combining canopy, ravine, water, and tree density to ask does this park give the city ecological respite? Each park sits somewhere on both. The chart below puts all 3,273 of them in one frame and refuses to collapse the answer.
Why this comparison matters
Park planning in North American cities tends to choose one tradition or the other. The Olmsted-and-after tradition reads parks as scenic relief. The city is the problem, the park is the cure, and the more it looks like uncultivated nature the better. The Jacobs tradition reads parks as functioning urban rooms: a park is a piece of the street network, alive because cafés and homes and schools press against its edges. Both traditions are correct about specific parks and wrong about others. They produce different policies, and they produce different cities.
The Atlas refuses to pick. Forcing a single Vitality score over all 3,273 parks would push ravines and parkettes toward the same number using different proxies, and the ranking would quietly favour whichever tradition won the weighting argument. Plotting the two axes against each other forces the question into the open: which kind of park is this, and is this what this neighbourhood needs?
Where Jacobs actually stood on parks
From The Death and Life of Great American Cities (1961), chapter 5.
Jacobs was specific about parks in a way that is often paraphrased away. She did not think parks were inherently good. She thought neighbourhood parks were created by the surrounding urban form, not the other way around. The same patch of green could be a treasured living room or a feared deadzone depending entirely on what pressed up against it.
“Conventionally, neighborhood parks or parklike open spaces are considered boons conferred on the deprived populations of cities. Let us turn this thought around, and consider city parks deprived places that need the boon of life and appreciation conferred on them.”
What earned a park its “life and appreciation,” in her telling, was four things, and the urban-integration metric is deliberately echoing them. She wanted intricacy (reasons to wander, things to see at the next bench), centering (a clear primary place inside the park), sun (not the constant tower shadow that came to define post-1970 plazas), and enclosure(a continuous mid-rise frontage looking down on the park, the source of her famous “eyes on the street” argument applied to green space).
Jacobs was openly skeptical of parks built without those four conditions. She called the wilderness-style green sea in a tower-in-the-park scheme a vacuum, and she was, by today’s standards, sometimes too quick to write off purely natural areas as beside the point. The Wilderness axis is here partly as a correction to that. Toronto has real ravine ecology, real urban canopy, and real cooling consequences, and a Jacobsian reading alone would miss why the ravine system matters at all.
For methodology and weighting see /methodology; for the Toronto-specific argument see the ravine paradox essay.
The Toronto Park Catalogue
Every Toronto park on the urban / natural plane. Colour = typology. Highlighted dots (outlined): Trinity Bellwoods, High Park, Nathan Phillips Square, Allan Gardens, Underpass Park, Grange Park, David Crombie Park. Hover any dot for details.
What the chart is showing us
Three things jump out the moment you stop reading dots one at a time and look at the cloud as a whole.
1. The Toronto Park Catalogue is bimodal. 12% of parks land in the urban-strong / nature-weak quadrant; another 34% land in the nature-strong / urban-weak quadrant; the centre is sparse. The city has been producing two parallel streams of green space (dense neighbourhood parkettes inside the streetwall, and a vast ravine and waterfront system outside it) and relatively little that does both. That isn’t accident. It mirrors how the city itself was built.
2. Genuine hybrids are rare. Only 220 parks (7%) score above 50 on both axes. The ones that do (Trinity Bellwoods, Christie Pits, Riverdale Park East, Allan Gardens) are the parks Torontonians most often name when asked their favourite. Whatever the model is picking up, it correlates with the parks people already love.
3. The bottom-left quadrant is full, and that’s informative too. 47% of the Toronto Park Catalogue scores below 50 on both axes. Most of these are placeholder parcels (hydro corridors, watercourse polylines, unprogrammed ravine slivers, transfer-station woodlots) that show up in the dataset as “parks” but don’t function as them. Surfacing this category honestly is more useful than hiding it: a quarter of the Toronto Park Catalogue is doing neither job, and that’s a real finding about how Toronto inventories its green space.
The model does not bell-curve these axes. The asymmetry in the cloud is the Toronto Park Catalogue’s actual asymmetry. See /methodology → Why scores are not bell-curved.
How the 3,273 parks divide
How to read each quadrant
Parks that work as urban rooms and as ecological respite. These are the rarest and the most-cited: Trinity Bellwoods, Christie Pits, Riverdale East. They emerged from neighbourhoods that happened to have mid-rise frontage and mature canopy at the same time. Almost none of them are new.
Strong urban integration, thin canopy. Civic squares, parkettes, market lanes. They reach people but offer little shade. These are the parks Jacobs would defend; they’re also the parks at most heat-vulnerability risk as Toronto summers lengthen.
Strong canopy / ravine / water, weak street life. The ravine system, urban forest fragments, parts of the islands. These parks deliver cooling, biodiversity, and mental respite, and many are quite hard to reach. The natural-comfort score is high but the connectivity score is low, and that gap is often by design.
Weak on both. Mostly placeholder parcels: hydro corridors, watercourse polylines, transfer-station woodlots, “City Wide Open Space” leftovers that the dataset classifies as parks but that don’t function as either an urban room or an ecological retreat. A few are real parks where the edge has decayed; those are the interesting ones.
Where each highlight sits
The seven dots called out on the chart, ranked by urban integration.
| Park | Typology | Urban | Comfort | Quadrant |
|---|---|---|---|---|
| David Crombie Park | Corridor / Linear Park | 75 | 59 | Balanced Hybrid |
| Grange Park | Neighbourhood Park | 75 | 42 | Urban Social |
| Trinity Bellwoods Park | Neighbourhood Park | 71 | 70 | Balanced Hybrid |
| Nathan Phillips Square | Civic Square | 62 | 32 | Urban Social |
| High Park | Ravine / Naturalized Park | 59 | 76 | Balanced Hybrid |
| Allan Gardens | Neighbourhood Park | 57 | 37 | Urban Social |
| Underpass Park | Athletic / Recreation Park | 44 | 26 | Underperforming |
Neither axis is the answer
A neighbourhood with only Urban Social parks bakes in July; a neighbourhood with only Ecological Retreats has nowhere to drop in after running an errand. The right question is rarely which park is best. It’s what mix does this resident reach in a 10-minute walk, and the chart can’t answer that on its own. That’s why the Atlas exposes both axes plus the Toronto Park Catalogue’s typology and cluster structure, so planning conversations can frame parks as a portfolio rather than a ranking.
Jacobs would have agreed with that framing. She wrote a chapter about parks and a chapter about the “great blight of dullness” a few pages apart, and the through-line was that no single design move fixes a city. What works is the right thing in the right place. The chart above is one attempt to make that visible at the scale of the Toronto Park Catalogue.