電磁垂直剖面探測浮標如何在低風速下量測海表面波浪頻譜

Abstract

本研究旨在評估電磁垂直剖面探測浮標（Electromagnetic Autonomous Profiling Explorer, EM-APEX floats）於低風速海域（約 5 ～ 8 m/s）進行海表面波浪量測的可行性與準確性。本研究提出的重心偏心旋轉校正法（Center-of-Gravity Eccentricity Rotation Correction, COG-ERC），可以有效修正儀器重量分布不均造成垂直加速度在量測時產生的偏差，並抑制垂直能量頻譜（E_Z (f)）於低頻區段（0.07<f≤0.1 Hz）的能量異常現象，大幅提升 EM-APEX floats 在方向波譜（E(f,θ)）、示性波高（h_s）、峰值頻率（f_p）與主波方向（θ_p）的估算品質。在 2 具浮標（編號 f9467 與編號 f9474）的資料應用結果顯示，該方法分別使低頻能量降低約 0.804 m^2/Hz 與 4.91 m^2/Hz。 為了驗證 EM-APEX floats 在波浪觀測上的表現，本研究將其估算的波浪統計參數與 NTU1 海氣象浮標上的 SVS - 603 波浪計，以及新海研 1 號研究船上的船用標準 X 波段波浪雷達進行跨平臺比對。結果顯示，當平臺間的距離 < 30 km 時，EM-APEX floats 與船用標準 X 波段波浪雷達在三項波浪統計參數的均方根誤差（root mean square error, RMSE）呈現：主波方向偏差（∆θ_p）為 29.29°、示性波高偏差（∆h_s）為 0.35 m、峰值頻率偏差（∆f_p）為 0.010 Hz，與 Beckman (2022) 以及 Lopez 等人 (2019) 的跨平臺研究結果相當，反映 EM-APEX floats 在低風速海況下的波浪觀測結果具有良好的可靠性。然而，當平臺間的距離擴大至 30 ～ 60 km時，∆θ_p 可能 > 90°；距離增加至 60 km 以上時，∆h_s 亦可能 > 0.4 m，顯示波浪場的空間非均勻特性隨平臺間的距離增加而放大，並且使跨平臺比對結果的參考性明顯下降。 EM-APEX floats 並非專為波浪量測所設計的平臺，然而，本研究證實其於低風速海域仍能穩定地提供具物理意義的波浪資訊。若進一步結合其多感測器整合能力與半拉格朗日（semi-Lagrangian）的漂流觀測特性，EM-APEX floats 即能彌補傳統波浪浮標與船載雷達系統在空間覆蓋與水文剖面量測的限制，展現其在海氣交互作用研究、多尺度波浪結構分析與海洋動力探勘的應用潛力，為一項具前景的輔助型觀測平臺。

Electromagnetic Autonomous Profiling Explorer (EM-APEX floats) provide simultaneous measurements of horizontal velocity from electromagnetic sensing and vertical acceleration from inertial sensors, offering a potential approach for estimating surface-wave properties under low-wind conditions. However, wave reconstruction can be biased by the float’s internal mass asymmetry, which introduces a rotation-dependent offset in the vertical acceleration and artificially elevates low-frequency energy in the vertical one-dimensional spectrum (E_Z (f)). To address this issue, we develop a Center-of-Gravity Eccentricity Rotation Correction (COG-ERC) method, which quantifies the systematic acceleration bias as a function of the float’s rotation angle and removes it through bin-averaging method (5°bins). Applying COG-ERC to two EM-APEX floats (f9467 and f9474) effectively suppresses the anomalous low-frequency energy in E_Z (f), reducing the mean energy in the 0.07 < f ≤ 0.10 Hz band by 0.804 m^2/Hz and 4.91 m^2/Hz, respectively. As a result, the reconstructed directional spectra (E(f,θ)) and key wave parameters—significant wave height (h_s), peak frequency (f_p), and dominant wave direction (θ_p)—are substantially improved under low-wind conditions (~5 – 8 m/s). The reliability of EM-APEX wave observations is further assessed through cross-platform comparisons with the SVS-603 wave sensor on the NTU1 buoy and a standard shipborne X-band wave radar onboard NOR1. When the separation distance between platforms is < 30 km, EM-APEX and the X-band radar show good agreement, with root mean square error values of 29.99°for ∆θ_p, 0.35 m for ∆h_s, and 0.010 Hz for ∆f_p, consistent with previous validation studies. In contrast, the comparability degrades as the separation distance increases (30 – 60 km and > 60 km), reflecting the growing influence of spatial wave-field heterogeneity on multi-platform comparisons. Although EM-APEX floats were not originally designed for wave measurements, our results demonstrate that—with appropriate correction—these floats can provide stable and physically meaningful wave information under low-wind conditions. Combined with their multi-sensor capability and semi-Lagrangian sampling characteristics, EM-APEX floats can complement conventional wave buoys and shipborne radar systems by improving spatial coverage while simultaneously resolving hydrographic profiles, highlighting their potential for air–sea interaction studies and multi-scale wave dynamics.