研究目的
Investigating the use of Landsat 5 data to detect surface changes caused by underground nuclear explosions (UNEs) for the purpose of localizing such events and supporting the Comprehensive Nuclear Test-Ban Treaty (CTBT) on-site inspections.
研究成果
Multispectral change detection using Landsat 5 data can effectively identify regions of interest potentially associated with underground nuclear explosions, supporting CTBT on-site inspections. The method demonstrated high sensitivity and improved separability between natural and anomalous changes, though its applicability is limited by environmental and data quality factors.
研究不足
The technique is less effective in areas with deciduous trees, variable atmospheric conditions, or snow. Requires sub-pixel ortho-rectification and datasets with minimal seasonal variability for optimal performance.
1：Experimental Design and Method Selection:
Utilized Landsat 5 data for change detection analysis, employing multivariate statistical analysis techniques including Mahalanobis distance and k-means clustering.
2：Sample Selection and Data Sources:
Selected Landsat 5 imagery from before and after the 1998 Indian and Pakistani nuclear tests, focusing on desert regions with minimal seasonal variability.
3：List of Experimental Equipment and Materials:
Landsat 5 satellite imagery, USGS Global Visualization Viewer (GLOVIS) for data access, and Google Earth for high-resolution imagery comparison.
4：Experimental Procedures and Operational Workflow:
Differenced pre- and post-event images, applied Mahalanobis distance for change detection, and used k-means clustering to analyze spectral characteristics of changed regions.
5：Data Analysis Methods:
Statistical analysis of temporal trends and data variability on a pixel-by-pixel basis, with emphasis on distinguishing natural from anomalous changes.
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
