Fusing spectral and structural airborne imagery data to detect escalating canopy dieback in a vulnerable eucalyptus population

Source
DataCite
Published
2026
Country
Australia
License
Creative Commons Attribution 4.0 International
Themes
Remote Sensing & Satellites, Forests, Climate & Temperature

Get the data from DataCite

About this dataset

Climate‑driven forest dieback is emerging as a major threat to biodiversity, yet early detection of canopy decline remains challenging when relying on single‑sensor remote‑sensing approaches. Here, we introduce an approach to investigate canopy dieback fusing high‑resolution aerial hyperspectral data (indicators of greenness) with LiDAR datasets (structural measures of biomass) to detect subtle changes in canopy condition that may not be captured by either dataset alone. Hyperspectral and LiDAR variables were integrated into a single raster stack to investigate a genetically unique Eucalyptus macroryhncha population in South Australia. Pixel‑based classification of canopy health into healthy, moderately healthy, and unhealthy classes achieved high accuracy for two surveys (Kappa = 0.88 in 2022; 0.81 in 2025). Between these surveys, environmental conditions shifted from post-drought...

How the data was collected

Climate‑driven forest dieback is emerging as a major threat to biodiversity, yet early detection of canopy decline remains challenging when relying on single‑sensor remote‑sensing approaches.

Browse sustainability datasets on SusanHub