利用衛星測高與GOCE大地水準面計算時變近海表面地轉流

Calculating Time-variant Near-surface Geostrophic Currents Using Satellite Altimetry and GOCE-determined geoid

基於海洋環流對於氣候穩定與人類生活的高度重要性以及其和天然災害之間的潛在關聯，長久以來提高海洋環流的求定精度以及持續性監測其變化一直是備受重視的課題。本研究採用Bosch and Savcenko（2010）所提出之軌跡法（Profileapproach），結合多顆測高衛星之沿軌跡海表面高度（Sea Surface Height, SSH）與GOCE第四代時域法解（The 4th generation of GOCE time-wise solution, GOCE-TIM4）重力場模型計算得之大地水準面，求得絕對海水面動力地形（Absolute Dynamic Topography, ADT），以減少因有限階數之GOCE大地水準面與SSH間空間解析力差異所產生的遺漏誤差（Omission error）對ADT之影響，配合由實測水文資料求得之相對動力地形（Relative Dynamic Topography, RDT）解算出全球海水各層地轉流流速。研究結果以熱帶大氣與海洋陣列／跨洋三角形浮標網（Tropical Atmosphere Ocean array/TRIangle Trans-Ocean buoy Network, TAO/TRITON）與熱帶大西洋預報與研究錨定浮標陣列（Prediction and Research Moored Array in the Atlantic, PIRATA）計27個錨定站之實測流速資料作驗證，結果指出使用軌跡法所求得之地轉流流速與實測資料間存在9~11cm/s之均方根（Root Mean Square, RMS），優於傳統逐點法（Pointwise approach）的12~16cm/s；而與傳統之逐點法相比，軌跡法約在60~80%之測站可得到更好的解算成果，且平均改善率（Rate Of Improvement, ROI）最高可達約30%，反映出軌跡法更能改善因大地水準面與SSH空間解析力差異造成之遺漏誤差對於地轉流流速計算之負面影響。 Due to the high importance of ocean circulations to climate stability, human life, and potential connection with natural disasters, the enhancement of determining accuracy and continuous monitoring of ocean circulation variations have been highly respected issues over the past centuries. The research adopts profile approach to process along-track Sea Surface Heights (SSHs) from multiple satellite altimetry and satellite-only geoid model (determined from satellite-only gravity field model - the 4th generation GOCE time-wise solution, GOCE-TIM4) to ease the negative impacts of omission errors to Absolute Dynamic Topography (ADT) resulted from the discrepancies in spatial resolution between truncated GOCE geoid model and SSHs. Resulting ADT was further combined with Relative Dynamic Topography (RDT) derived from in-situ hydrographic data to determine mesoscale geostrophic current velocities at different depth layers globally. Results were validated by in-situ current meter observations at 27 moored stations obtained from Tropical Atmosphere Ocean array/TRIangle Trans-Ocean buoy Network (TAO/TRITON) and Prediction and Research Moored Array in the Atlantic (PIRATA), indicating that profile approach gives Root Mean Square (RMS) differences of them at 9~11 cm/s better than those of 12~16 cm/s given by conventional pointwise approach. Profile approach also outperforms conventional pointwise approach at around 60~80% of stations with an average Rate Of Improvement (ROI) up to 30%, showing that profile approach better solves the detrimental impacts resulted from omission errors.

Due to the high importance of ocean circulations to climate stability, human life, and potential connection with natural disasters, the enhancement of determining accuracy and continuous monitoring of ocean circulation variations have been highly respected issues over the past centuries. The research adopts profile approach to process along-track Sea Surface Heights (SSHs) from multiple satellite altimetry and satellite-only geoid model (determined from satellite-only gravity field model - the 4th generation GOCE time-wise solution, GOCE-TIM4) to ease the negative impacts of omission errors to Absolute Dynamic Topography (ADT) resulted from the discrepancies in spatial resolution between truncated GOCE geoid model and SSHs. Resulting ADT was further combined with Relative Dynamic Topography (RDT) derived from in-situ hydrographic data to determine mesoscale geostrophic current velocities at different depth layers globally. Results were validated by in-situ current meter observations at 27 moored stations obtained from Tropical Atmosphere Ocean array/TRIangle Trans-Ocean buoy Network (TAO/TRITON) and Prediction and Research Moored Array in the Atlantic (PIRATA), indicating that profile approach gives Root Mean Square (RMS) differences of them at 9~11 cm/s better than those of 12~16 cm/s given by conventional pointwise approach. Profile approach also outperforms conventional pointwise approach at around 60~80% of stations with an average Rate Of Improvement (ROI) up to 30%, showing that profile approach better solves the detrimental impacts resulted from omission errors.