As a research assistant for Assistant Professor of Environmental Studies Heather Kropp, I’ve learned to look at black-and-white aerial images to examine the city of Utica. The dark, slanting blacks suggest shadows. Those shadows help me see the block buildings or the trees’ textured grays. When I need to find the streets of 1950s Utica, I use present-day images as a guide. Each tree, building, or street I see gets outlined. Each outline gets plugged into the system. That system, operated and created by Kropp, will produce historical spatial data sets and historical maps of Utica, which will be accessible to researchers and community members interested in how the Utica environment has changed over decades.
“There’s a wealth of historical imagery out there,” Kropp said. “I find it really interesting to look at the geography of the past and compare that to the geography of the now, especially to see the ways that plants shape and change our environment.”
Historical imagery presents a unique challenge because, although we can make out an image’s features with our eyes, it is difficult for the computer to recognize and quantify them. However, with newer data science approaches, we can “teach” the system how to do it. Each time I outline a tree and that outline gets plugged into the system, the system develops a better understanding of what a tree looks like. It can then use this information to make accurate predictions about which pixels make up a tree, allowing the system to quantify different images’ features and create strong spatial data sets. These steps in the data production process are known as workflow. This workflow could offer insight for other researchers working with historical imagery.
“I want to build a workflow that is completely transparent and reproducible,” Kropp said. “I want anyone to be able to adapt it to their own codes and needs.”
When Kropp began this project in the fall, she chose to use this workflow on a project in Utica because she wanted to have a local focus and fill a need. As a small city, Utica lacks a lot of the spatial data that exists for other cities. The lack of data limits the research that can be done on Utica, its history, and its present-day complexities.
As I work on outlining various features of Utica, I am motivated most by the future research this project could create. Knowing where trees are helps us to determine who has access to green spaces. Being able to track the decrease of trees and increase of pavement in various areas helps us determine if the change was equitable or if it may have targeted certain populations. These potential points of research are grounded in environmental justice, and they can help us learn more about the equity of Utica’s urban development.
Kropp hopes to finish this project this summer. She suspects that her own research will build off of the project results, offering new opportunities to students interested in local environmental justice.