| 摘要: | 地磁場變化受地球內部核心及太陽風等多種因素的影響，包含來自電離層及磁層共同效應、潮汐效應、海水導電度差異、火山活動、磁暴事件或其他地球外部效應影響，雖大多皆屬極小尺度之變化，但對於理解地磁變化之物理意義仍具有重要影響力，故連續性之監測將有利於獲取相關資訊，以利深入探究。中央氣象署管理之全臺12座連續三分量磁力測站，本研究使用主成份分析（Principal Component Analysis, PCA）處理測站獲取之連續性時間序列資料，嘗試自長時間段數據中提取有意義之資訊，解析地磁場變化之主要模式及其物理機制。本研究首先探討傳統頻譜分析（Spectral Analysis）之優劣，發現其因單純針對訊號頻率，難以有效區分有效訊號與雜訊，因而改用主成份分析以同時考量時間與空間的變化。藉此地磁場變化可被分解為多個獨立的主成份 ，每個主成份可能對應於特定的地球物理現象，例如日變化（Sq variation）、電離層效應或地質構造之影響。此外，分析各測站對不同主成份之貢獻程度，能獲取地磁變化之空間分佈特徵。另外將進一步探討主成份隨時間變化的模式，並嘗試找出與潮汐、磁暴及地震等活動之潛在關聯性。成果將有助於提升對地磁變化之理解，透過主成份分析，能更精細解析地磁場的多種變異機制，可為地震前兆監測建立新的數據分析框架，並為將來地球物理探測與研究提供更深入的見解。;Previous studies have suggested that geomagnetic field variations are influenced by a variety of factors, including the Earth’s internal core processes and external forces such as solar wind. These variations arise from combined effects of the ionosphere and magnetosphere, tidal forces, differences in seawater conductivity, volcanic activity, geomagnetic storms, and other extraterrestrial influences. While most of these changes are small in magnitude, they play a critical role in understanding the physical mechanisms behind geomagnetic variations. Therefore, continuous monitoring is essential to acquire valuable information for in-depth investigation.
The Central Weather Administration of has established 12 continuous three-component geomagnetic observatories across the island. This study applies Principal Component Analysis (PCA) to the long-term time series data collected from these stations, aiming to extract meaningful information and identify major patterns and mechanisms driving geomagnetic field changes.
This study will begin by examining the advantages and limitations of traditional spectral analysis. Although effective in identifying frequency components, it struggles to distinguish between meaningful signals and noise. In contrast, PCA considers both temporal and spatial variations, allowing the decomposition of geomagnetic variations into independent principal components. Each component may correspond to specific geophysical phenomena such as Sq (solar quiet) variations, ionospheric effects, or geological structures.
By analyzing the contribution of each station to different principal components, spatial patterns in geomagnetic variability can be inferred. Furthermore, we explore the temporal evolution of these components to identify potential correlations with tides, magnetic storms, and seismic activities.
The results of this study will contribute to a deeper understanding of geomagnetic variations. By applying PCA, it becomes possible to more precisely resolve the multiple mechanisms driving changes in the geomagnetic field. This approach may offer a useful data analysis framework for monitoring potential earthquake precursors and provide valuable insights for future geophysical exploration and research. |
