摘要
MODIS影像是浒苔业务化卫星遥感监测的重要数据源,但其空间分辨率(250m)较低,混合像元效应导致传统的NDVI(Normalized Difference Vegetation Index)阈值法浒苔提取误差较大。为解决此问题,本文以准同步的较高分辨率的HJ-1CCD影像浒苔提取结果为基准,建立基于MODIS NDVI与浒苔像元丰度(浒苔覆盖面积占像元面积百分比)关系的浒苔信息提取模型,利用该模型可提取每个像元的丰度(即浒苔覆盖率),从而提高整景影像浒苔覆盖面积提取精度。该模型可为浒苔灾害等级的确定以及防灾减灾应急决策提供支持。
MODIS images are valuable data source generated from operational satellite remote sensing for monitoring Enteromorphain the Yellow Sea.However,there are large errors inenteromorpha coverage derived on base of NDVI Threshold Method,because of mixed pixels in the coarse resolution（250m）MODIS images.To solve the problem,this paper,using the near real-time HJ-1CCD images as reference,the＂NDVI-abundance＂decomposition of mixed pixel model was established.The developed model can significantly improve the extracted Enteromorphainformation.As the method presented here can extract more accurate information,it is speculated to play a significant role in copying with Enteromorphadisasters.
出处
《海洋学报》
CAS
CSCD
北大核心
2015年第7期123-131,共9页
基金
国家海洋局青年海洋科学基金项目(2012405)
海洋公益性行业科研专项经费项目(201205010
2013418025-2)
国家海洋局第一海洋研究所基本科研业务费专项资金项目(GY0214T03)
