摘要
针对随着全球气候变化、人口剧增和社会经济的发展,湿地生态系统及其生物多样性面临前所未有的威胁和干扰的现状,该文通过利用若尔盖湿地国家级自然保护区植被类型样点提取的长时间序列中分辨率成像光谱仪归一化植被指数同一像元值数据,采用非度量多维尺度分类方法,分析其提取植被类型的可靠性,并对植被生长期变化特征进行了研究。研究结果表明:长时间序列原始与Savitzky-Golay重构的归一化植被指数数据按植被类型分类结果显著,应力系数分别是0.07和0.04;保护区植被归一化植被指数值与降水量、温度存在中度、高度线性相关;保护区沼泽植被生长期呈延长趋势,草甸和沙地植被生长期在时间上有向后推移的趋势。
With the global climate change, population explosion and the development of social econo- my, the wetlands ecosystem and biodiversity are facing an unprecedented threat and interference. Therefore, it is very important to monitor the change of wetland vegetation. The paper analyzed the reliability of vegetation types and the growing variation of vegetation types using non-metric multidimensional scaling based on the data extracted from long-time series MODIS NDVI of the Zoige plateau wetland national nature reserve. The results showed that the result of vegetation classification was remarkable by using long time series original and Savitzky-Golay reconstructed NDVI data. Stress was 0.07 and 0. 04, respectively. Vegetation and precipitation were significant linear moderate correlation. Vegetation and temperature were significant linear correlated. The growing period of marsh vegetation extended, and meadow and sand vegetations growing period went backwards in terms of time.
出处
《测绘科学》
CSCD
北大核心
2015年第6期110-114,共5页
Science of Surveying and Mapping
基金
国家自然科学基金资助项目(41071265
41372340)
高等学校博士学科点专项科研基金课题项目(20105122110006)
国家科技支撑计划项目(2012BAH28B00)
四川省地理国情监测工程技术研究中心资助项目(GC201405)
关键词
非度量多维尺度法
植被类型
可靠性
生长期
non-metric MDS
Savitzky-Golay
vegetation type
reliability
growing variation
