During the period of 1996-1999, an exploratory analysis on in situ hyperspectral data of conifer species was conducted. Based on in situ hyperspectral data measured using a spectrometer, PSD1000, the hyperspectral data of conifer species important to northern California, USA, were analyzed. Using spectral derivative technique, artificial neural network algorithm, and multivariate statistical analysis, such as analysis of variance, linear discriminant analysis, and principal component analysis methods, the spectra of six conifer species: Douglas-fire, Giant-sequoia, Incense-cedar, Ponderosa-pine, Sugar-pine and White-fir, whose spectral properties are very similar, were recognized; differences of spectra at different needle ages were analyzed; and foliage nutrient status (total nitrogen, total phosphorus and total potassium) of Giant-sequoia was estimated and predicted with in situ hyperspectral data. This study led to four journal papers published.
Please go back my main homepage