You cannot register for this meeting

Upcoming Office Hours: Dec 6: Enhancing North American LANDFIRE Landscape Classifications with Machine Learning & Automating LCP File Generation in the Cloud using Sentinel 1 & 2 in South America Speaker 1: Dr. Inga La Puma Since its start, LANDFIRE has used machine learning, particularly supervised classifications, with vegetation plots as training data to classify area pixels. These classifications inform fire behavior fuel models. With advancements in data and cloud computing, there's potential for deeper, more frequent landscape classifications using sophisticated algorithms. Researchers are exploring these advanced techniques with LANDFIRE. Speaker 2: Dr. Andrea Markos Mapping Fire Behavior Fuel Models’ complexity lies in ambiguous classes and their focus on “expected fire behavior,” not Land Cover. Clear training data from US-based LANDFIRE layers allow transfer learning globally, automating the LCP file generation applying machine learning to current satellite data.