研究目的
Investigating the use of Landsat 5 data to detect surface changes caused by underground nuclear explosions (UNEs) and to develop change detection techniques specific to the use of satellite data to support a Comprehensive Nuclear Test-Ban Treaty (CTBT) on-site inspection.
研究成果
The study demonstrates that multispectral data can be used to find regions of interest potentially associated with a UNE. Change detection is a sensitive technique to identify those regions of interest, provided there is useful training data. The Landsat 5 data were useful for identifying potential regions of interest before an on-site inspection has begun.
研究不足
The technique is not likely to be useful in areas with deciduous trees, variable atmospheric conditions, or snow. Data ortho-rectification must also be sub-pixel. The method's effectiveness depends on the scene characteristics and the nature of the test.
1：Experimental Design and Method Selection:
The study employs multivariate statistical analysis techniques, including principle components analysis (PCA) and Mahalanobis distance, on temporally differenced spectral bands from Landsat 5 data.
2：Sample Selection and Data Sources:
Landsat 5 imagery of the 1998 Indian and Pakistani nuclear test sites, downloaded from the USGS Global Visualization Viewer (GLOVIS).
3：List of Experimental Equipment and Materials:
Landsat 5 satellite data, calibrated to reflectance units.
4：Experimental Procedures and Operational Workflow:
Analysis involves differencing pre-event and post-event images, applying statistical techniques to identify changes, and using k-means clustering to categorize changes.
5：Data Analysis Methods:
Statistical analysis of the data to understand temporal trends and data variability on a pixel-by-pixel basis, with the aim of distinguishing natural changes from those indicative of a UNE.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
