Chlorophyll content estimation of Boreal conifers using hyperspectral remote sensing

摘要

This investigation quantitatively links physiologically based estimators of forest stand condition, such as chlorophyll concentration, to hyperspectral observations of Jack Pine (Pinus Banksiana), a dominant Boreal Forest species. Between June and September of 2001, four Intensive Field Campaigns (IFC) of data collection were conducted over the forested areas near Sudbury, Ontario, Canada. Using the CASI sensor, data were collected, in the visible and near infrared domain, over eight selected Jack Pine sites. Supplementing the airborne campaigns was simultaneous on-site collection of foliage samples for laboratory spectral and chemical measurements. The study first linked needle-level reflectance and pigment content through the inversions of leaf level radiative transfer models such as PROSPECT. Next, the red-edge index (R750/710), was scaled up to the canopy level through the use of canopy models and infinite reflectance calculations, which simulate the canopy as an optically thick vegetation medium. However, for the relatively open and clumped jack pine stands such a simple approach requires careful validation due to the confounding effects of the open canopy structure. Accordingly, the analysis has focused on high spatial resolution CASI imagery (1 meter) for which tree crowns, shadows, and open (sun-lit) understory can be identified visually and approaches can be examined for validity and effects. Effectively eliminating these confounding variables will permit the generation of predictive needle pigment content maps for forest condition assessment.