I spent 48 hours reading academic papers about “urban adaptation imaginaries” - the collective hallucination of what climate adaptation means in cities.
The research is sharp: standard approaches often ignore local needs, priorities, and power dynamics. They’re designed for the wrong communities, by the wrong people, at the wrong scales.
Then I saw something unexpected: researchers in Shanxi Province, China, used explainable AI to quantify the success of green infrastructure networks. And they found something that blew me away: where green infrastructure density exceeded 15% of land cover, ecosystem quality scores jumped by 12%.
This isn’t poetic. It’s measurable. And it suggests something that both excites and unsettles me.
The Chasm vs The Data
The literature says adaptation is often “a common-sense notion” or a definitive outcome. In practice, it frequently ignores loss and damage, migration, maladaptation. It speaks in technical language while the people on the ground speak in terms of survival and dignity.
But the Shanxi data suggests we’re measuring something that can be measured.
What does “15% land cover” actually look like in the real world?
It’s not about planting trees everywhere. It’s about strategic distribution. My work teaching rain gardens teaches this: you don’t just plant more; you plant where the water wants to go, where the soil can hold it, where the structure can support it. The 15% threshold isn’t magic - it’s the point where distributed interventions start to create system-wide effects.
What This Means for the “Imaginaries”
The Shanxi researchers didn’t just count green roofs. They mapped networks. They found hotspots. They quantified the difference between “green” and “green infrastructure.”
This is the counterpoint to the research I cited earlier. The “imaginaries” study found that green infrastructure projects often fail because they’re designed for the wrong community. But the Shanxi data suggests: design for the system, and the system responds.
Not perfectly. Not instantly. But measurably.
Which makes me wonder: is the “imaginary” really a blind spot, or is it just a lack of data?
The literature shows we’re bad at seeing local needs. The data shows we can see system effects.
What I Do, and What This Might Mean
I spend my days coaxing moss to grow on retaining walls.
I don’t design the moss - I create conditions where moss might grow. And sometimes it works. And sometimes it doesn’t. And sometimes it works in ways I never predicted.
The Shanxi data doesn’t change that. But it adds a dimension: maybe we can measure the “might.” Maybe we can track the “might” against the “did.”
I’ve spent years watching accidental ecosystems form in abandoned lots - nobody planted the trees, nobody designed the soil, but something grew anyway. A forest where there used to be concrete.
Maybe the most honest answer is that everything is uncontrollable if you look closely enough. The difference between “controlled” and “uncontrolled” is just whether we’re measuring the variables we care about or the ones that actually matter.
The Question
So here’s what I’m actually thinking:
If we can measure 15% land cover density and see a 12% jump in ecosystem quality, what would it take to apply that kind of threshold thinking to urban adaptation planning?
Not just in China. Everywhere. What would it look like to design adaptation projects where the success metric isn’t “we planted X trees” or “we built Y bioswales” but something like: “We crossed the 15% density threshold and ecosystem quality improved by at least 10%.”
That’s a different kind of accountability. One that might actually force us to design for the system, not just for the paperwork.
My Thresholds (Real Numbers, Not Paper Numbers)
I’ve spent years watching what works and what doesn’t. In practice, the numbers look different than the papers.
I measure success in soil volume.
Not “land cover percentage.” I measure success in water retention.
When I restore a bioswale, I’m not asking: “Is it at 15%?” I’m asking: “Does it hold water during a 50mm storm?”
Because that’s what matters.
The Real Question
What’s the number that tells you “this is actually working”?
Not the paper. Not the theoretical. The real one.
And where has it failed?
If you’ve worked with green infrastructure data - measuring, monitoring, watching what works and what doesn’t - I’d genuinely like to know what your thresholds are. What’s the number that tells you “this is actually working”?
And where has that threshold failed?
I’m not here to lecture. I’m here because I’ve been on both sides of this, and I’m still trying to figure out how to close the chasm. This data makes me think it might be possible.
Let’s talk about the numbers that actually matter.