February 13, 2025, 1-1:45 PM ET
Hosts: UConn CLEAR
Speakers: Chandi Witharana, Assistant Professor, Durga Joshi, PhD Student, Harshana Wedegedara, PhD Student, University of Connecticut, Department of Natural Resources & the Environment
The increasing availability of multimodal remote sensing datasets offers tremendous opportunities to map landscape features and track their changes across space and time. This 45-minute webinar will showcase several new statewide geospatial datasets generated from the latest 2023 leaf-off LiDAR and imagery collection. Connecticut hosts terabytes of high-resolution remote sensing data in online repositories, such as CTECO, ranging from 1930s grayscale aerial photographs to modern multispectral aerial images and high-density 3D LiDAR point clouds. The latest 2023 statewide leaf-off LiDAR survey comprises over 700 billion individual points at a density of roughly 14 points/m², complemented by a concurrent multispectral aerial imagery collection exceeding two trillion image pixels at 8 cm spatial resolution. Integrating LiDAR with aerial imagery enables accurate feature extraction and classification, making it virtually possible to map every urban tree, utility pole, and powerline segment in the state. However, traditional algorithms often struggle under the sheer volume and complexity of these large datasets, underscoring the need for high-throughput, GeoAI-powered analysis pipelines that can operate at scale. This webinar will showcase geospatial products including tree canopy height, plant area index, and foliage height diversity, and will discuss how these products contribute to downstream analysis and synthesis efforts, such as urban tree equity, forest carbon modelling, and roadside vegetation management.