期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2000
卷号：XXXIII Part B7(/1-4)
页码：768-775
出版社：Copernicus Publications
摘要：The use of temporal records of vegetation properties oriented to understand the influence of weather and climate ecosystems has been a traditional research technique. Vegetation is an extremely sensitive indicator of changes in many factors of the environmental behaviour such as climate, soil and land management. Since the availability of long records of satellite observations has become easier to researchers the challenge to monitor the vegetation behaviour and to correlate it with the climate process has influenced recent remote sensing studies. A well-known parameter accepted by researchers of the remote sensing community for the study of vegetation is the normalised difference vegetation index (NDVI). The methodologies aiming at the study of vegetation temporal dynamics attempt to measure the amplitude, periodicity, or any other temporal variation of the vegetation index. One of the most appropriate tools for the study of the dynamics of any system is Fourier transform. It became a more attractive tool particularly after the development of the fast Fourier transform (FFT) algorithm. In the last decade a new tool, known as wavelet transform (WT) has made its appearance for the study of stationary and non-stationary signals. The purpose of the present study is to point out some particular features of the wavelet transform that can be used to characterise the dynamical behaviour of the NDVI time series. The final goal of this work is to study the dynamical behaviour of the vegetation cover in some regions of South America by using monthly NDVI data and wavelet transform (WT) techniques as a mathematical tool
关键词：Wavelets; NDVI; AVHRR; Remote Sensing; Time Series
