DF seminar: Deep Learning for Wildfire Monitoring with Earth Observation Time Series
We welcome Puzhao Zhang who is a Digital Futures post-doc researcher at the Division of Geoinformatics at KTH, working with Professor Yifang Ban.
Date and time: 3 December, 12-1 pm
Speaker: Puzhao Zhang
Title: Deep Learning for Wildfire Monitoring with Earth Observation Time Series
Meeting ID: 674 3268 2790
Abstract: Across Sweden and around the globe, wildfires are growing in both intensity and frequency, which is highly correlated with the rising global temperature. Wildfires emit massive amounts of carbon dioxide and other pollutants that could affect regional and even global climate, in turn, climate changes contribute to more and bigger wildfires. To reduce the impact of wildfires on the climate and the damages from such destructive wildfires, it is urgent to develop robust methods for monitoring wildfires over vast remote areas across the globe. Satellite-based Earth Observation (EO) systems provide a cost-effective way to monitor the changes on the earth’s surface caused by wildfires.
To effectively monitor changes using satellite imagery, it is critical to develop Artificial Intelligence (AI)-based methods to automatically extract timely information on wildfire progressions from available EO data. Capable of penetrating clouds and imaging at day and night, Synthetic Aperture Radar (SAR) data has the unique advantages of monitoring on-going wildfires that are often accompanied with heavy smoke which may block optical observations. In this talk, I will present some of our recent research on SAR and fusion of SAR and optical data for wildfire progression monitoring with deep learning.
Bio: Puzhao Zhang is a Digital Futures post-doc researcher at the Division of Geoinformatics at KTH, working with Professor Yifang Ban. He received his PhD degree in Pattern Recognition and Intelligent Systems from Xidian University, China at the end of 2019. He has been working on wildfire monitoring as a joint PhD student at KTH since Oct., 2017.
His research objective is to apply advanced machine learning techniques on spatio-temporal EO big data to help to tackle the environmental challenges our society is facing, and his research interests include satellite imagery analysis, change detection, machine learning, and spatio-temporal modelling.