A summary of Asian vegetation using annual vegetation dynamic indicators

Abstract

Time series Advanced Very High Resolution Radiometer (AVHRR) data of eastern Asia for the period April to October 1992 have been analyzed to produce statistically independent annual indicators of vegetation development. Monthly AVHRR data composites used in the analysis were generated by USGS EROS Data Center using the maximum Normalized Difference Vegetation Index (NDVI) approach to produce a cloud free representation of eastern Asia. The time series data were analyzed to determine whether a phenologically based transformation could assist analysis of vegetation dynamics and understanding of landscape features. The reflective channels of each monthly composite were transformed to Tasseled Cap greenness and brightness. Comparison of the sun angle corrected greenness feature to NDVI favored use of greenness in the analysis. Using principal components methods, a linear transformation of the lime series greenness was discovered. The transformation describes ninety percent of scene greenness variation in four variables. These variables include cumulative greenness (the sum of greenness for each date), summer greenness flux (the contrast of summer biomass productivity to the rest of the year), early season greenness flux (a measure of early season biomass productivity), and periodic greenness flux (biomass productivity attributable to multiple cropping). The transformation enables effective visualization of vegetation dynamics in a single image, capturing dominant biological patterns. Imagery formed using annual vegetation dynamic indicators illustrates the potential of employing the transformation to interpret land cover, and identify natural temporal dynamics as well as those that are a result of human intervention, like multiple cropping.