Discriminating Tsuga canadensis Hemlock Forest Defoliation Using Remotely Sensed Change Detection.

Area of Science:

• Forest Ecology
• Remote Sensing
• Invasive Species Management

Background:

• Eastern hemlock (Tsuga Canadensis) populations are declining in North America.
• Infestation by the hemlock woolly adelgid (Adelges isugue), an invasive insect, causes significant defoliation and tree mortality.
• Effective monitoring is crucial for understanding the spatial and temporal extent of this forest health crisis.

Purpose of the Study:

• To investigate the application of Landsat satellite imagery and change-detection techniques for monitoring hemlock forest health.
• To develop and validate a remote sensing methodology for assessing hemlock defoliation.
• To establish a baseline for tracking hemlock health trends over time.

Main Methods:

• Utilized Landsat satellite imagery and change-detection algorithms to analyze spectral reflectance changes in hemlock canopies.
• Applied atmospheric correction methods to ensure data accuracy over time.
• Related changes in the Normalized Difference Vegetation Index (NDVI) to ground-truthed levels of hemlock defoliation.
• Developed a classification system to predict four distinct classes of hemlock forest health.

Main Results:

• Successfully predicted hemlock forest health classes with 82% accuracy across heterogeneous landscapes.
• Identified spectral reflectance changes in near-infrared and red wavelengths correlating with defoliation.
• Established a methodology to monitor temporal and spatial patterns of hemlock defoliation over a decade.

Conclusions:

• Landsat-based remote sensing is a viable and accurate tool for monitoring the health of eastern hemlock forests.
• The developed methodology provides a scalable approach for regional forest health assessments.
• Expansion of this monitoring approach to the entire Mid-Atlantic region is underway.