基于MODIS植被指数的太湖蓝藻信息提取方法研究  

Extraction methods of cyanobacteria bloom in Lake Tai based on MODIS vegetation index

作　　者：李亚春[1] 孙佳丽 谢志清[3] 谢小萍[1] 

机构地区：[1]江苏省气象台,南京210008 [2]江苏省气候中心,南京210008 [3]江苏省气象科学研究所,南京210008

出　　处：《气象科学》2011年第6期737-741,共5页Journal of the Meteorological Sciences

基　　金：中国气象局新技术推广项目(CMAT2007M30)

摘　　要：有效地提取蓝藻水华信息对分析蓝藻动态分布有重要意义,而卫星遥感技术是进行太湖水质监测与保护的措施之一。本文以2007年7月25日Terra/MODIS数据为主要数据源,用比值植被指数(Irv)、归一化植被指数(Indv)和增强型植被指数(Iev),研究提取太湖蓝藻信息。结果表明:植被指数可以有效提取遥感影像中的蓝藻水华信息,其中Indv是应用效果较好的植被指数之一;在掩膜处理后,用Indv提取了太湖蓝藻的面积分布信息,效果较好;此外,选取Indv为测试变量,利用决策树分类法,有效地把蓝藻水华高浓度覆盖区、中浓度覆盖区和轻浓度覆盖区分开来,为准确掌握太湖蓝藻发生、发展趋势和发生程度提供可靠信息。Abstract It is very important to extract the information of cyanobacteria bloom-forming in order to analyze the distribution of cyanobacteria bloom, and remote sensing is one of measures of monitoring and protecting water quality in Lake Tai. In this paper, Terra/MODIS data of July 25, 2007 were selected as a main data source. The information of cyanobacteria bloom was extracted by three kinds of MODIS vegetation index（Ⅰrv,Ⅰndv ,Ⅰev）. The results showed that these vegetation index methods were able to getcyanobacteria information from the image, and that Ⅰndvis one of the methods which have better applica- tion effect. The area information of cyanobacteria bloom was also extracted by Ⅰndv after mask treatment. Then the areas covered by high, middle or low of cyanobacteria bloom was able to be separated by build- ing decision tree with Ⅰndv as the test variable. Therefore, the present study provides technical reference for long-term monitoring cyanobacteria bloom-forming in Lake Tai.

关 键 词：蓝藻 植被指数 决策树 

分 类 号：X87[环境科学与工程—环境工程]