Mapping spatio-temporal variation in Douglas-fir (Pseudotsuga menziesii) foliar biochemistry

Abstract

A high resolution imaging spectroscopy study was conducted, which investigated the spatial and temporal diversity of biochemistry in Douglas-fir (Pseudotsuga menziesii) stands of southern Vancouver Island, Canada. An AVIRIS dataset with a nominal spatial resolution of 4m was acquired coincident with sample acquisition that measured organic and inorganic chemistry. An AISA dataset was colledted for the same study site. Partial least-squares regression was applied and chemistry was modeled from reflectance. Nitrogen, chlorophyll (a, b) were selected as contributing to reflectance responses. Chemistry estimation achieved R2 values ≥ 0.93 for both datasets. The models applied to the hyperspectral imagery yielded two temporally discrete datasets of biochemistry. Change detection analysis was conducted with attention to chlorophyll a/b for its ability to indentify regions of forest disturbance and stress. Factors affecting biochemistry were investigated with reference to laboratory data (management strategy and maturity).