基于多时相遥感影像的亚像元级地表水变化监测研究

项目名称： 基于多时相遥感影像的亚像元级地表水变化监测研究

项目编号： No.41501460

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 天文学、地球科学

项目作者： 黄昌

作者单位： 西北大学

项目金额： 20万元

中文摘要： 对地表水的动态监测往往需要基于多时相的遥感影像进行，而鉴于中低分辨率影像在时间分辨率上的优势，它们在此类研究中的作用不容忽视。但是，它们相对较低的空间分辨率又使得监测的精度因混合像元问题而大受影响。基于单个时相遥感影像的混合像元分解和亚像元制图是当前处理这类问题的常规方法，但这类方法在效率、精度、监测结果的稳定性上存在不足。本项目以NPP-VIIRS数据为例，探索基于多时相遥感影像在亚像元尺度上监测地表水动态变化的方法。该方法结合变化检测方法中的光谱变化向量分析法与线性光谱混合模型，探索混合像元内水体丰度变化的监测方法，然后通过结合多时相的水体分布信息改进亚像元制图算法，从而最终实现地表水降尺度监测的完整流程。本研究结合野外调查、高分辨率遥感影像解译，对所提出方法的精度和适用性进行论证。研究成果将丰富地表水遥感降尺度监测的方法体系，促进中低分辨率遥感数据在地表水体动态监测研究中的应用。

中文关键词： 混合像元分解；亚像元制图；超分辨率制图；地表水；可见光-近红外遥感

英文摘要： Monitoring surface water dynamic is generally based on multi-temporal remotely sensed images. Medium- to low-resolution images, due to their high temporal resolution, have been playing a significant role in this research area. However, mixed pixel issue is an inevitable problem that will affect the accuracy of monitoring. A common way to deal with this issue is pixel unmixing and sub-pixel mapping based on single-temporal remotely sensed image, which has limitations in efficiency, accuracy and consistency. This project tries to develop a methodology for monitoring surface water dynamic at sub-pixel scale based on multi-temporal remote sensing images. NPP-VIIRS imagery has been chosen as experimental data. This study combines the spectral Change Vector Analysis method and Linear Spectral Mixture Model to detect water fraction change within mixed pixels. After that, it integrates multi-temporal surface water distribution information into sub-pixel mapping algorithm. These two procedures compose a complete process of surface water downscaling monitoring. This study will then testify the accuracy and applicability of the proposed methodology based on a combination of field survey and interpretation on high-resolution remote sensing images. The research results will enrich current methodology of surface water downscaling monitoring. They will also promote the application of medium- to low-resolution images in surface water dynamic monitoring.

英文关键词： pixel unmixing;sub-pixel mapping;super-resolution mapping;surface water;visible-near infrared remote sensing