Satellite remote sensing is particularly useful for understanding large scale terrestrial carbon cycle. Canopy optical properties depending on leaves and trunks characteristics are used in estimating leaf area index (LAI) from satellite observations. Because a canopy reflectance respond to a leaf phenology, it is necessary to consider a plant phenology for estimating temporal variations in LAI accurately. In this study, we estimated spatial and temporal variations in LAI from a new canopy radiative transfer model with a plant phenology scheme.

First, we evaluated temporal variations in the LAI by using the new model for each plant functional type in Japan for several years. We used optical properties of leaves, trunks and forest floors and a canopy reflectance measured with a spectral radiometer as model inputs. The model estimations showed reasonable seasonal patterns at the validation sites. Then, we calculated spatial and temporal variations in LAI from the MODIS reflectance products in East Asia. As a result of regional scale analysis, we could understand that misclassifications of land cover map contributed to the decreased accuracy of estimations in a comparison of the seasonal variations at point scale.