Assessing Leaf Area and Canopy Biochemistry of Florida Pine Plantations Using Remote Sensing

Abstract

Leaf area index (LAI) and the biochemical makeup of forest canopies are critical determinants of carbon (C) gain by forests. However, both variables may be highly unstable in space and time and high spatial resolution measurements made on the ground are difficult and expensive to obtain. We estimated LAI and canopy biochemistry using remotely sensed imagery and data from a field site in north central Florida. Ground measurements indicated a high seasonal fluctuation in LAI, and annual variation of approximately 10%, correlated with the Normalized Difference Vegetation Index (NDVI) derived from Thematic Mapper (TM) imagery. Repeated fertilization affected concentrations of nitrogen (N) and chlorophyll in the pine foliage, but had little effect on concentrations of water, lignin or cellulose. Biochemical differences among samples of whole fresh pine needles were related to reflectance properties determined in the laboratory using a portable spectroradiometer. We then explored coupling laboratory spectral measurements of foliar biochemistry with field spectra by analyzing the signal-to-noise ratio (SNR) of the Airborne Visible Infrared Imaging Spectrometer (AVIRIS). Results indicate that the AVIRIS provides data with an SNR barely sufficient to estimate foliar biochemistry; maximum SNRs for slash pine are suggested. Finally, using actual AVIRIS data and simultaneously obtained field samples, stepwise regression of corrected imagery indicated that three wavebands accounted for 94% of the spectral variation, all in the spectral region of the reflectance red edge. These collective results indicate the feasibility of parameterizing process-level models of primary productivity of P. elliottii stands using remotely sensed data.