GIS, Remote Sensing and Machine Learning: Data Integration to Support the Management of Coastal Island Ecosystems.

ABSTRACT: Islands are important components of many coastal areas around the world; however, by virtue of their geographical isolation, the state of these ecosystems is often poorly known. To address the knowledge gap for the province of Nova Scotia, Canada, geographic information systems (GIS), remote sensing (RS), and machine learning (ML) were used to examine the status of nearly 4000 islands. We classified islands topographically and determined, based on 1 m resolution LiDAR, that approximately 70% are <2 m average elevation and highly vulnerable to partial or complete flooding under near-term regimes of sea level rise and storm surge potential. Vegetation cover was strongly related to topographic class, with higher, more steeply-sided islands having more tree cover and less sand, rock, and wetla

SUBMITTER: Lieske DJ 

PROVIDER: S-EPMC12457486 | biostudies-literature | 2025 Oct

REPOSITORIES: biostudies-literature