This past July, I was invited to present our Urban Environment and Social Inclusion Index (UESI) at the World Economic Forum’s (WEF) Annual Meeting of the New Champions (AMNC ’19) in Dalian, China as part of the Global Situation Space forum. This space allows for researchers to utilize the Earthtime Platform, developed by Carnegie Mellon’s CREATE Lab, to tell data-rich stories on a range of themes, including the topic I was tasked to address on urbanization and growing fragility within cities.
The pilot UESI, launched in December 2018 at the United Nations Climate Conference in Katowice, included 35 cities from around the world. In time for the WEF AMNC ’19 presentation, the Data-Driven Lab team worked tirelessly to add more than 130 new cities to the Index. This expansion allowed us to verify whether the trends in the initial findings held, at least for the geospatial indicators (i.e., air pollution, urban heat island, tree cover, and transportation) that were easy to calculate after lab members TC Chakraborty and Diego Manya automated their calculation using Google Earth Engine.
We are in the process of finalizing these updated results and will release them on the Data-Driven Lab’s urban site soon. I will also give an encore presentation at the WEF’s Sustainable Development Impact Summit in New York City from Sept. 23-24, where participants will be able to comment and express viewpoints on the need for sustainable innovation to address the growing inequality of environmental goods and harms in cities throughout the world.
I also had the opportunity to speak on a panel, “Smart Solutions: Detecting Pollution,” with David Lehrer, Chief Executive Officer, Conatix, United Kingdom. We discussed the potential for advances in machine learning, cloud computing, and real-time sensing to improve on pollution detection. While there are many promising gains through machine learning in this space, including productivity gains, improved outcomes, cost reduction, discovering unknowns through big-data mining, David warned about the fact that data are still subject to human interference and manipulation. Polluters can still game models or poison datasets. Errors can be large and unknowable – at times when it could be too late. People can easily drown in the sea of big data and lose the ability to derive actionable insight.
I presented some practical, real-life examples from our own work at the Data-Driven Lab where we see large-scale data, machine learning, and systematic analysis leading to some new tools for decisionmakers to have unprecedented access to information that could improve environmental decision-making. The UESI is one example of such a tool – TC Chakraborty’s global urban heat island (UHI) explorer is making data on the UHI and land-surface temperature accessible to more than 9,000 urban areas around the world.