Multi-Sensor Validation and Scaling of Machine Learning Models for Contrail Detection and Flight Attribution using Integrated Ground-Based and Satellite Observations

This proposed research project aims to tackle the significant climate impact of aviation contrails by creating a high-quality “ground truth” dataset to improve how we detect and predict them. By extending Western University’s extensive Global Meteor Network of ground cameras to European ground-based camera systems and improving current machine learning models, the project will help solve the “flight matching” problem – accurately identifying which specific flights create warming contrails. This collaboration offers a powerful synergy: Western University benefits by transforming its astronomical camera network into a world-leading tool for climate monitoring, while TU Delft gains the essential ground-based machine learning algorithms to extend the use of their current cameras. Ultimately, this partnership equips researchers with the specialized skills and validated models necessary to lead global initiatives in reducing aviation’s contribution to climate change.

Faculty Supervisor:

Denis Vida

Student:

Partner:

Delft University of Technology

Discipline:

Physics

Sector:

Education

University:

The University of Western Ontario

Program: