基于多源遥感数据的藻华信息提取研究

Information Extraction of Algal Bloom Water Body Based on Multi-Source Remote Sensing Data

将同期的ENVISAT ASAR图像和MODIS图像进行配准,利用经过大气校正后的MODIS光学图像,根据分类决策树对图像EVI进行分类,借此对太湖的藻华和非藻华区域进行划分,并以此为据在ASAR图像上研究两者的V/V后向散射特性,使用阈值法对藻华范围进行提取,并将结果与MODIS图像上得到的藻华范围进行对比。结果表明,通过ASAR图像与MODIS图像得到的藻华范围较好的吻合,说明在一定条件下,ASAR图像可以显示出藻华。最后,本文还从机理上探讨了风速对藻华成像的影响,分析了光学和微波图像提取藻华范围不完全一致的原因以及利用微波数据监测藻华的局限性。

ASAR and MODIS images are registered together.After applying atmosphere correction to MOSIS image,we can use an empirical model to retrieve chlorophyll-a concentration,then the algal bloom can be identified.Based on this classification,the V/V backscattering characteristics of algal in ASAR image has been examined and the water body containing algal can be extracted using threshold classification method.Then,a comparison has been made on this classification between ASAR and MODIS image and the result matches well.So a conclusion can be made that the algal bloom can be shown on the SAR image under certain conditions.In the way of mechanism,this paper discusses the influence of wind on algal bloom and tries to analyze the differences of the classification results as well as the shortcomings of SAR image when applied to monitor the algal bloom.This work suggests a new way to monitor water quality of Taihu lake using SAR technology.