Columbia Researcher Designs Data Visualization of Carbon Footprints
A Columbia researcher affiliated with the Data Science Institute and the Earth Institute has created a data-visualization tool that shows the carbon footprints of hundreds of consumer products. The tool makes it easy for everyone to explore the products’ carbon emission levels and the various strategies companies are employing to reduce emissions.
The visualization tool, called Carbon Catalogue, breaks down the carbon footprint of a product during its entire life cycle, illustrating the carbon it emits during the raw material, manufacturing and later downstream phases. The data show that several companies have made vast improvements in reducing their products’ emissions. Some have instituted sustainable practices such as reducing packaging for food and beverage products, while others replaced fossil fuel with bio energy or lowered the energy consumption of computers.
“This free tool can serve as an inspiration for other companies to reduce their products’ emissions, especially since smaller carbon footprints often correlate with reduced production costs,” says Christoph Meinrenken, an associate research scientist at The Earth Institute and chief data scientist at CoClear, an environmental analytics firm.
The visualization looks like a wheel with color-coded spokes, each representing a consumer product such as a cell phone, a car, or a pair of jeans. When a user hovers over a spoke, a pop-up box appears with a summary of a product’s life-cycle data, including improvements companies made to reduce its carbon emissions. Working with CoClear, Meinrenken analyzed carbon emissions data for 866 products made by 145 companies from 28 countries. The companies voluntarily submitted the data to CDP (formerly the Carbon Disclosure Project), a nonprofit that asks companies to complete detailed questionnaires about their products’ emissions data. The researchers used life-cycle data submitted to CDP on products from 2013-2017 to create the visualization, whose menu allows viewers to search by company, industry or year.
“We designed Carbon Catalogue, which is free and open to all, to enhance awareness of carbon sources and ways to reduce them,” says Meinrenken, who also belongs to the Earth Institute’s Research Program on Sustainability Policy and Management. “The tool gives consumers and CDP-member companies like Nestle, Bloomberg and Dell an interactive platform to explore products’ carbon footprints.”
Meinrenken and his CoClear colleagues found that upwards of 75 percent of a product’s overall emissions were not generated during the manufacturing process, as one might expect, but rather during the supply-chain phase, such as when raw materials are acquired or when consumers later use the products. The food and beverage industry, for instance, which relies heavily on agricultural inputs, had one of the highest upstream emissions, while the IT industry, with its energy-consuming screens and computers, had one of the highest downstream emissions.
“Only by comparing upstream and downstream emissions can you find the hotspots that tell you where you can eliminate the most carbon,” says Meinrenken. “We hope that product managers will use the visualization to see how their products’ carbon footprints fare in comparison to comparable products made by competitors and find ways to reduce emissions.”
With the advent of big data, companies can now mine the minutiae of their supply chains for insights that benefit both their business and the environment. This capability also comes at a time when consumers, customers, employees, and investors are demanding details about a product’s effects on the environment. And companies that responded to the demand by performing life-cycle analysis on products’ emissions from “cradle to grave,” reduced their carbon footprints at twice the rate of companies that only tracked a product’s total emissions, the researchers found.
“Clicking your way through the tool and exploring the many ingenious ways companies found to make cleaner products is quite heartening,” says Meinrenken. “I see this as a continuation validation of long-term efforts by many of my colleagues in the corporate and public sector to make life-cycle analysis more scalable and wide spread.”
Though the data visualization tool is easy to use, the work that went into making it was difficult. Conducting a life-cycle analysis of a product’s carbon footprint is labor-intensive and the data companies submitted to CDP needed “a lot of curating.” Companies submitted their product emission data in myriad formats with varying levels of detail, and Meinrenken had to process the reams of data. To evaluate a simple consumer product such as a soda can, for instance, companies had to collect data on “the masses of three packaging materials and five ingredients, transportation distances of all materials sent to the plant, total amounts for four types of energy as well as transportation distances to stores and refrigeration times in stores and at home, after which all materials and activities have to be paired with respective carbon emission factors, bringing the count of data inputs to about 100 for a single product.”
To gain insight from the data, the researchers designed algorithms to mine the data submitted to CDP and convert every product’s footprints into a universal set of metrics that allowed for apples-to-apples comparisons across products and sectors. With these metrics in hand, Meinrenken and CoClear worked closely with the Data Science Institute to refine the visualization tool. During a Sustainability Hackathonhosted by DSI, data science students developed code and created visualizations that Meinrenken and CoClear’s Erika Whillas later used for the final design. Spry Group, a New York-based web development agency also helped them design Carbon Catalogue.
Researchers at Coclear are currently adding 2016 and 2017 data to the visualization model, and they say some positive trends in the carbon-reporting ecosystem are already clear: 20 percent of companies that pursued carbon reduction in 2016-2017 did so specifically after receiving data requests from their supply-chain partners, who were already reporting to CDP.
CoClear was formed in 2012 by two Columbia graduates: Sally Paridis, CoClear’s chief executive officer, has a master’s degree in sustainability management from Columbia’s School of Professional Studies, while Erika Whillas, CoClear’s chief product officer, has a master’s in public administration from the School of International and Public Affairs. A third Columbia graduate, Daniel Chen, joined as director of data analytics after finishing his master’s in earth and environmental engineering at Columbia Engineering.
When Paridis was a student at Columbia, she took a life cycle assessment class with Meinrenken, who is also an adjunct professor in the Department of Earth and Environmental Engineering, Columbia Engineering. Even then, in 2011, she talked to him about designing a publicly available platform on which to list the life-cycle analysis of carbon footprints of products. At the time, though, companies did not keep sufficiently detailed emissions data on their products, so she put her idea aside. With CoClear, she has since focused on helping companies make product assessment faster and cheaper, using some of the patented methods Meinrenken developed at Columbia. And now, eight years later, her idea for a carbon-footprint platform has come to fruition in this innovative data visualization tool, one that has promise to reduce carbon emissions and mitigate global warming.
“I’m dedicated to finding sustainable ways to reduce carbon emissions and improve our environment, which I see as the most pressing challenge of the 21st century as well as a business opportunity,” says Paridis. “This Carbon Catalogue of product emissions can be used by a global network of customers, consumers and policy makers to investigate supply chains and unlock opportunities to reduce carbon emissions. When people realize the journey is starting for most companies, my hope is that they’ll feel no shame about getting started—no guilt—just ready to take action.”
— Robert Florida, Data Science Institute