Predicting European cities' climate mitigation performance using machine learning

Area of Science:

• Environmental science
• Urban planning
• Climate change mitigation

Background:

• Cities are increasingly recognized as key players in global climate action.
• A significant challenge in assessing urban climate performance is the scarcity of reliable emissions data.
• Evaluating city-level mitigation efforts requires robust and scalable methodologies.

Purpose of the Study:

• To develop and apply a machine learning approach for evaluating climate mitigation performance in European cities.
• To assess trends in carbon dioxide (CO2) emissions for local administrative areas across Europe from 2001 to 2018.
• To investigate the relationship between participation in climate initiatives, data reporting, and emissions reduction performance.

Main Methods:

• Utilized a machine learning model integrating publicly available environmental and socio-economic data with self-reported emissions.
• Applied the model to nearly all local administrative areas in Europe for the period 2001-2018.
• Predicted annual CO2 emissions to analyze city-scale mitigation trends.

Main Results:

• European cities engaged in transnational climate initiatives demonstrated a likely decrease in CO2 emissions since 2001.
• Over half of these cities appear to have met their 2020 emissions reduction targets.
• Cities that reported emissions data exhibited greater emissions reductions compared to those that did not.

Conclusions:

• The developed machine learning approach offers a scalable and replicable method for assessing city-level climate mitigation performance.
• Data reporting by cities is positively correlated with successful emissions reductions.
• Further research and data transparency are crucial for accurately tracking and enhancing urban climate action.