Abstract:

Leaf Area Index(LAI) is a key structural and functional biophysical variable in the study of ecological environment research, also, an important path of energy exchange of terrestrial vegetation. Through the study on radiative transfer theory of vegetation from multi-angular, hyper-spectral data, LAI is estimated by bidirectional reflectance distribution function(BRDF). According to discrete anisotropic radiative transfer(DART) model, the zenith angle and azimuth are discretized to 61 direction combinations, look-up-table(LUT) of LAI inversion which is suitable for multi-angular and  hyper-spectral data is constructed. Then, PROBA/CHRIS data in Changbai Mountain area is selected, pre-processed, and matched with LUT based on least squares principle, LAI in the research area is retrieved. Subsequently, the sensitive factors of LAI to BRDF, of 470 nm and 700 nm bands, in the 61 discrete directions, are evaluated. With LAI values increasing, BRDF value of 490 nm wavelength increases firstly and decreases successively, and BRDF value of 700 nm wavelength decreases; BRDF value decreases when zenith angle increases, and reaches minimum near the zenith.

Key words: remote sensing, hyperspectral imagery, DART model, leaf area index, CHRIS data, look-up-table