Skip to main content


Angel Hsu presented “Seeking the Signal Through the Noise: The Data-Driven EnviroLab.” She provided an overview of the Data-Driven EnviroLab, an interdisciplinary and international group of researchers, scientists, programmers, and visual designers, discussing global climate action from a data science approach. Hsu detailed two of the group’s projects:   Global Climate Action - Data Science Approaches The Data-Driven EnviroLab seeks to understand the net global effect of those currently acting on climate change, including governments, businesses, hybrid coalitions, and private actors, to understand the gap between needed actions and pledged actions. This is not an easy task, however. Hsu conveyed that this type of data science is largely “janitorial work.” It involves cleaning and putting together datasets, using machine learning and language processing tools to pull information from pdfs, and taking advantage of fuzzy matching algorithms to compile and harmonize everything into one dataset. To scale this work, EnviroLab is working with partners to develop an internet for climate change, where people do not have to submit their information to a centralized repository but can connect about these topics. Additionally, the Data-Driven EnviroLab is working with Google to develop a machine infrastructure to understand emissions inventories using geospatial data and machine learning.   Urban Environment and Social Inclusion Index Geospatial data can be used to address environmental questions, particularly questions involving the way that the environment and environmental pollution and climate change issues do not affect everyone uniformly. The Urban Environment and Social Inclusion Index (UESI) project uses satellite remote sensing and geospatial data to measure the differences in air pollution, urban heat, and other factors that affect access to clean air and clean water. For example, Hsu discussed variations in tree cover for neighborhoods of different financial means in Los Angeles, citing that higher income areas have significantly more tree cover, which has huge implications for health, social inclusion, air pollution exposure, and heat exposure. To fully understand the nature of these problems, however, she discusses a persistent need for improved data, which they have begun to collect. For example, to get a sense of interurban heat variability, EnviroLab had students carry sensors to produce data, like air temperature and humidity, about Chapel Hill. This information allowed the group to build a model (map) and predict heat exposure across a variety of locations in Chapel Hill.   Click here to view the talk on YouTube.


Angel Hsu, Assistant Professor; Director, Data-Driven EnviroLab

Department: Public Policy and Environment, Ecology and Energy Program | Faculty Profile

Featured on: October 27, 2022 (Event Page)

Session Title: Unearthing Environmental Impact (Event Recap

Tools, Information, and Resources:

  • Data Driven EnviroLab: The Data-Driven EnviroLab uses innovative data analytics to distill signals from large-scale and unconventional datasets and develop policy solutions to contemporary environmental problems. Working with scholars and policymakers across the globe, the Data-Driven Lab strives to strengthen environmental policy at all levels.
    • Natural Language Processing: A collaboration with Arboretica to develop automated data extraction algorithms for unstructured data.
    • Data Engineering: A collaboration with OpenEarth Foundation and OS-Climate to develop federated data commons to connect and showcase existing climate data. The platform aims to automate as much of the data ingestion pipelines and metadata management as possible.
  • Urban Environment and Social Inclusion Index (UESI): The Urban Environment and Social Inclusion Index (UESI) is a first-of-its-kind tool that leverages high-resolution, large-scale data to reveal how cities perform at the intersection of environment and social equity. 
  • Data Driven Heat Hack: The Data-Driven EnviroLab and Museum of Life and Science HeatHack 2022 Data Science Collabathon brought together coders, programmers, data visualization whizzes, and environmental enthusiasts to help us discover actionable insights, develop engaging data visuals, and create action plans to address the growing and intensifying challenge of climate change and urban heat in the Triangle.
  • National Academy of Sciences: Greenhouse Gas Emissions Information for Decision Making: A Framework Going Forward (2022): Climate change, driven by increases in human-produced greenhouse gasses and particles (collectively referred to as GHGs), is the most serious environmental issue facing society. This report examines existing and emerging approaches to generate and evaluate GHG emissions information at global to local scales.