Conventional industrial and urban ecology models treat cities as a spatially isolated system, ‘black boxing’ the geographies of production where resource extraction occurs. Centered on domestic aggregate supply chains, this work traces the movement of construction aggregate materials (limestone, gravel, and sand) from quarries to cement plants, concrete producers, and urban cores. 

By spatializing material flow networks, this work moves past the ‘black box’ of the urban metabolism, illuminates the distal ecological trade-offs and equity implications of an urbanizing planet, and identifies strategic intervention points for global land stewardship.

  Project GitHub

Concrete is one of the most carbon-intensive materials by volume; in NYC, buildings account for nearly 70% of the city's total GHG emissions. 

NYC's Buy Clean initiative incentivizes developers to invest in 'green concrete' to reduce these embodied emissions. This project uses a bottom-up Material Flow Analysis (MFA) to quantify concrete accumulation in new building construction across all five boroughs from 2012–2033.

This study asks a deceptively simple question: does a Buy Clean standard reduce upfront emissions in NYC's building sector, and who bears the burden of this decision?

We find that, while a Buy Clean policy cuts emissions intensity, it does not eliminate emissions growth and total emissions still rise due to construction volume growth.

Project GitHub

Agent-based modeling (ABM) is a bottom-up computational approach used for understanding the structure of collective behaviors based on hard-coded rules. 

This study examines the use of large language models (LLMs) in ABM to simulate epidemic simulations, guided by a deceptively simple question: does the behavior of infected agents change when agents make LLM-informed decisions?

Project GitHub