应用R/S方法分析NDVI时间序列

针对1982--2001年月合成PALNDVI时间序列，应用R／S方法计算中国陆地范围内的赫斯特指数。结果表明：全国范围内的赫斯特指数普遍大于0．5，意味着NDVI时间序列不是马尔科夫序列，而是存在明显的趋势性；中国陆地范围内的赫斯特指数分布存在“西高东低”的特点，说明西部NDVI时间序列表现出更强的持久性；通过分析赫斯特指数分布与陆地覆盖类型的关系发现，在同一生态地区内，裸地的赫斯特指数最高，与其它覆盖类型差异明显。陆地覆盖类型之外的自然条件和人为因素对赫斯特指数空间分布的影响有待进一步研究。

Application of R/S Method in Analyzing NDVI Time Series

NDVI time series data are valuable in researching various geo - phenomena on the earth surface. In this paper, the R/S method was the first time adopted to calculate the Hurst exponents of NOAA/NASA PAL monthly NDVI time series from 1982 to 2001 on Chinese terrestrial regions. Since 99.5 percent of Hurst exponents of the NDVI time series are greater than 0.5, it is obvious that NDVI time series are generally persistent,and a rising tendency in the past will lead to some increase in the future. Mostly higher Hurst exponents distribute in west China,which means that there is stronger persistence for NDVI time series in the west China than in the east. When land cover types are considered,bating land shows much higher Hurst exponents than other land- cover types. As NDVI is affected by various natural and human factors,the relations between such factors and Hurst exponents of the NDVI time seties are to he quantified in the coming research.