Classifying Forest Succession with LiDAR Data

LiDAR data have proven particularly useful for measuring a suite of forest structural attributes such as canopy height, basal area, and LAI. However, to date, the potential of LiDAR data to quantify forest successional stage remains largely untested. Our objective is to evaluate the use of an algorithmic modeling approach incorporating LiDAR data to classify forest succession across a structurally diverse, mixed-species forest in Northern Idaho. We focus upon using various LiDAR metrics to classify forest succession according to a succession classification scheme developed for the Inland Northwest. The final classification identified six classes of forest development (stand initiation, young multi-story, mature multi-story, understory reinitiation, and late successional) with an overall accuracy > 95%. The techniques presented herein can be easily applied to other forested ecosystems. Furthermore, the final forest succession classification could be used for many purposes including forest succession modeling and wildlife habitat assessment.