近地表剪力波速度（Shear-wave velocity， Vs）為工程地質與場址效應評估中之重要參數。多頻道表面波分析法（Multichannel Analysis of Surface Waves， MASW）因具備非破壞性與施測效率高等優點，已廣泛應用於淺層地盤調查。然而傳統 MASW 多仰賴離散式地震檢波器陣列，其空間取樣密度與長距離監測易受儀器數量與佈設條件限制。分布式聲學感測（Distributed Acoustic Sensing， DAS）可利用光纖作為連續式感測元件，提供高密度與長距離之波場觀測能力，但其資料品質仍受光纖耦合、佈設方式與場址環境影響，因此有必要進一步評估其應用於主動式與被動式分析之可行性。 本研究以 DAS 結合主動式與被動式分析為核心，於國立中央大學、國立陽明交通大學與新竹五峰鄉土場邊坡三處場址進行現地試驗。三處場址皆進行主動式落錘試驗與被動式環境背景噪訊記錄，其中國立中央大學國泰樹場址使用 Sintela ONYX Peta 與 Terra15 兩台 DAS 系統進行量測比較。主動式資料透過事件選取、訊號堆疊、頻散分析與剪力波速度反演，評估 DAS 取得表面波頻散資訊之能力；被動式資料則利用互相關函數與相位加權疊加重建虛擬表面波，以取得被動式頻散曲線並進行速度反演。 研究結果顯示，在適當資料處理與良好光纖耦合條件下，DAS 可記錄沿光纖測線傳播之表面波訊號，並取得可辨識之頻散能量分布。主動式資料因震源位置明確，較有利於取得穩定之頻散曲線；被動式資料則需透過長時間背景噪訊累積補充不同頻段之波場資訊，但其品質仍受記錄時間長度、噪訊來源分布與光纖耦合狀態影響。整體而言，主動式與被動式 DAS 表面波分析皆具有應用於近地表剪力波速度反演之潛力，兩者頻散曲線之比較與整合可作為工程場址評估與長期監測之參考。;Near-surface shear-wave velocity (Vs) is an important parameter in engineering geology and site-effect assessment. Multichannel Analysis of Surface Waves (MASW) has been widely applied to shallow subsurface investigations because of its non-destructive nature and high survey efficiency. However， conventional MASW generally relies on arrays of discrete geophones， and its spatial sampling density and applicability to long-distance monitoring are often limited by the number of instruments and field deployment conditions. Distributed Acoustic Sensing (DAS) utilizes optical fiber as a continuous sensing element， providing high-density and long-distance wavefield observations. Nevertheless， the quality of DAS data is still affected by fiber–ground coupling， deployment configuration， and site conditions. Therefore， further evaluation is required to assess the feasibility of applying DAS to both active- and passive-source surface-wave analyses. This study focuses on the integration of DAS with active- and passive-source surface-wave methods. Field experiments were conducted at three sites: National Central University， National Yang Ming Chiao Tung University， and an earth slope site in Wufeng Township， Hsinchu County. Active-source drop-weight tests and passive ambient-noise recordings were performed at all three sites. At the Cathay Tree site of National Central University， two DAS systems， the Sintela ONYX Peta and Terra15， were additionally deployed for comparative measurements. The active-source data were processed through event selection， signal stacking， dispersion analysis， and shear-wave velocity inversion to evaluate the capability of DAS to retrieve surface-wave dispersion information. For the passive-source data， cross-correlation functions and phase-weighted stacking were applied to reconstruct virtual surface waves， from which passive dispersion curves were extracted and used for velocity inversion. The results indicate that， with appropriate data processing and favorable fiber–ground coupling conditions， DAS is capable of recording surface waves propagating along the fiber-optic cable and producing identifiable dispersion-energy distributions. Active-source data are more favorable for obtaining stable dispersion curves because the source location and excitation time are well defined. Passive-source data， in contrast， can supplement wavefield information over different frequency ranges through the accumulation of long-duration ambient noise. However， the quality of the passive-source results remains dependent on the recording duration， noise-source distribution， and fiber–ground coupling conditions. Overall， both active- and passive-source DAS surface-wave analyses show potential for near-surface shear-wave velocity inversion. The comparison and integration of dispersion curves obtained from the two approaches may provide useful information for engineering site assessment and long-term monitoring.
