Urban Technology at University of Michigan week 114
Interview with Alex Kobald on the Tree Folio project
Planting trees along city streets is an ancient tactic gaining renewed attention as cities understand their treescape’s role in handling extreme heat and volatile weather, offering a way to mitigate the heat island effects that see temperatures rise in dense, paved urban centers. Here in Ann Arbor, the city has pledged to plant 10,000 new trees by 2030. Seoul’s tree plan includes the creation of “wind path forests” that will funnel cool, clean air into the city center. These are just two examples of the many cities that have committed to increasing their urban canopies.
Alongside these initiatives, there is a parallel trend to count and classify the trees already planted in a city, since it’s difficult to know where to plant new trees without understanding the existing streetscapes. Functionally, too, a city needs to understand where trees are, and what kind they are, in order to maintain them. So, how best to understand the current state of trees in a city?
Enter Tree Folio: NYC, described as “a digital simulation of each street tree in NYC and its local shading impact … to study the distribution of street tree benefits and inform more equitable and effective tree planting strategies.” Combining data from NYC’s 2015 tree census and a 2017 LiDAR scan, Tree Folio offers 3D models of each tree in the city. We spoke with Alex Kobald, co-director of the Design across Scales lab at Cornell University and lead for the Tree Folio project, to learn more.
💬 Hello! This is the newsletter of the Urban Technology program at University of Michigan, in which we explore the ways that technology can be harnessed to nurture and improve urban life. If you’re new here, try this short video of current students describing urban technology in their own words.
🌳 Street Trees as Infrastructure
CHARLIE KEENAN: Why trees? How did you get started working on this project?’
ALEX KOBALD: We started on this project as part of a discussion group organized by Cornell’s Urban Tech Hub examining how well NYC’s 3-1-1 system was working to maintain the city’s street trees. Early on, we were interested in how to ensure these maintenance requests were leading to the equitable maintenance of trees. As we worked on this further, we began to ask if trees and the benefits they provide were equitably accessible in the city in the first place, and if not, how we could begin to quantify a tree’s benefits and where those benefits are currently distributed. Since most of our team has design and urban planning backgrounds, we decided to focus on the shade benefits that trees provide, as those are the most directly tied to the physical shape and location of a tree and are so clearly related to some of the pressing concerns around heat island effects and extreme heat events. Our lab at the time was also working on a project with a nonprofit organization called Street Lab who had drawn our attention to the vulnerability of lower income communities to extreme heat. That collaboration began to inform our framing of Tree Folio as a way of treating the urban canopy as a distributed shading infrastructure, with all the questions of historic inequality and civic ownership that a word like “infrastructure” entails.
CHARLIE: Walk us through the datasets you are using for this project—how are they generated, and what are their strengths and weaknesses?
ALEX: Our source data is a 2017 LiDAR survey of NYC. That is the dataset that we use to extract the physical models of each tree canopy, and that physical model is used to calculate the shading benefits. We then use the 2015 NYC Street Tree census to check the accuracy of our predicted tree location add information about the tree species (if there is a street tree near the predicted tree location). Both datasets are publicly available and have some strengths and weaknesses. The LiDAR dataset is very high resolution and covers the entire city, allowing us to extract tree canopies that are public street trees, park trees, and trees on public land. Additionally, these scans are available for most cities and towns in the U.S., meaning that if our workflow works in NYC, we can deploy it in smaller, less resourced cities across the country or abroad. The source data is from aerial surveying using high frequency light pulses and measuring the amount of time it takes the pulse to return to the survey source. The end result is a file of points, with XYZ positions, and metadata including the type of return signal (one-of-one return, or one-of-multiple returns) and some preliminary classification.
The weakness is that this dataset is very large. There are approximately 1,200 tiles for NYC, and each tile is a file of about 1 GB and contains millions of points. There are ~1 million trees in NYC, each described by 2,000-11,000 points. This requires quite a bit of computational resources and some optimization to produce the models and shading metrics we are trying to extract. The Tree Census is a helpful dataset that has very accurate locations and species information for the city’s street trees. Street trees are only part of the urban canopy, though, so it falls short in describing access to shade from all types of trees in the city. Also, the Tree Census is conducted by human volunteers, making it very time-consuming to update.
CHARLIE: What are some of the insights you’ve had in designing a research tool that will be used by clients outside of the academy?
ALEX: We’ve just started to discuss this project with stakeholders, so these are still early days. So far, I’ve learned again about the abnormal position that NYC is in, in the sense that there are many organizations thinking about the urban canopy. They all point out that they really need more resources, but it is amazing to have diverse perspectives around tree canopies that include the Department of Parks and Recreation, the Mayor’s Office of Resilience, various park conservancies, and a whole host of neighborhood NGOs and advocacy groups that want to plant more trees and learn more about the benefits of the trees they already have. All these stakeholders have their own ideas and perspectives on what a “better” urban canopy would look like, which is helpful to discuss and think about.
CHARLIE: What differences would you expect to see if you applied the same methods to a smaller city, or to a city outside of the U.S.?
ALEX: I think some of the differences would clearly be in the local contexts we would be considering. NYC is very dense, and a lot of the buildings are taller than the trees they are next to. As a result, elements like street orientation and width have a huge impact on the amount of shade a tree provides. In other settings there could be different factors that are more important to maximizing a tree’s local impact. Additionally, in NYC we can add information about a tree’s species, which adds a lot of specificity to the isolated model and its local impact. Many cities and towns in the U.S. have tree census surveys, but not all, so losing that piece of information does make the individual models a bit more abstract.
CHARLIE: Who all is working on this project? If you were going to add another person to the team, what would their skills and superpowers be?
ALEX: Currently our team includes: Joe Ferdinando, formerly from the Cornell AAP M.Arch program and now doing a master’s in computer science, who is working on the shading simulation and visualization; Sarang Pramode from the Urban Tech Hub working on the canopy extraction and shading simulation; Jiahao Dong from the Urban Tech Hub working on visualization; myself as the project lead and coordinator; and Dr. Anthony Townsend as an advisor. The Urban Tech Hub has been a supporting and collaborating partner in this project from the beginning as well.
There are two types of people we’d love to add to the team. The first would be someone with skills in front-end web design and an interest in 3D visualization to really push the interactive communication aspect of this project as far as it can go. The second would be a subject matter expert on some of the other benefits trees can provide (evapotranspiration, air quality, carbon sequestration) and on public policies to evaluate where those benefits are most needed.
CHARLIE: What’s next for this project?
ALEX: We have two short-term goals. The first is to finish building the shading simulation and visualization tool, so that we can quantify the benefits of each tree in NYC and communicate those benefits in a clear and compelling way. The second is to have conversations with subject matter experts, policy makers, community-based advocacy groups, and other organizations who can expand our ideas on how this tool can be used to have real-world impacts. Our goal would be to have a tool that really matters to these stakeholders, and to then be able to deploy our work and the ways to use it effectively to cities around the country and abroad.
CHARLIE: And lastly, as we ask everyone we interview: what’s your favorite city, and why?
ALEX: I’m from Vancouver, British Columbia and have yet to go to a city I’ve liked more. It has some of the best parks in the world, lots of greenery, and really diverse neighborhoods and sights to see. I like the outdoors, so being within a short drive of mountains and glaciers is a big plus to me as well. One criticism is that it isn’t as walkable a city as you’d probably expect—and ironically, I think it could use a better strategy for the urban canopy.
These weeks: grinding the gears on double majors, electives, and other academic minutiae; call with Tongji; ordering tables for our first (!) studio; gearing up for admissions by making elaborate spreadsheets; autumn. 🏃