| 摘要: | 先前之研究指出，過去在檢視模式之風暴潮水位時，可能出現與觀測資料比對後，總水位扣除觀測值之殘差失準問題，其可能因素包含極端事件之影響、特定區域之地形特殊性、潮位站坐落於港口內部而非外海等，因此有必要單獨檢視純風暴潮水位與天文潮汐水位之影響。然而，在單獨檢視潮汐水位時，依然會出現水位數值之擺盪情形，顯示有非風暴潮之因素影響了水位之準確辨識。
為瞭解水位殘差之來源，本研究透過COMCOT-SS數值模式（Cornell Multi- grid Coupled of Tsunami Model - Storm Surge），並搭配TPXO8-atlas之天文潮模式作為邊界條件，藉由求解非線性之淺水波方程式，模擬多個現有潮位站在特定純潮汐期間之水位結果，並進行統計參數之收斂性分析，並發現潮汐水位之到時對總水位殘差影響之可能性。
在探討港灣構造對水位於時間上之影響，其中涉及多種變量，包含曼寧係數、水位之波高、潮位計相距距離、水深等，為建立參數之間之關聯性，本研究透過簡化港口地貌特徵與進行波速之無因次參數I_z之無因次化分析（I_z值之計算方式為：I_z=C_s/C_p，C_s為模擬波速，是藉由模式結果中，兩不同位置潮位計之距離與入射波到時所計算出；C_p為預估波速，由波速計算式：C_p=√gH（其中， g為重力加速度，H為水深）所計算出），紀錄模擬設置下港灣構造對於港口內外之水位時間差異性，觀察港灣構造之遮蔽效應對於水位模擬之波速之潛在影響，並於後續進行實際港口案例之水位分析，以分析不同變量對水位到時之影響程度。
研究結果指出，上述三個不同變量（波高h、曼寧係數n、港內水道之水深H）所計算出之無因次參數x_1 、x_2 、x_3所改寫之無因次參數皆有可能成為水位出現時間誤差之因素，並且曼寧係數n所計算出之無因次參數x_2與波速無因次參數I_z為負相關，而波高h所計算出之無因次參數x_1則與波速無因次參數I_z為正相關。最終，透過統計所有資料點進行線性回歸分析後，整理出經驗公式，並與實際港口之模擬案例比對後，確認該經驗公式確實可用於評估港內外潮位計之波速與時間差異。
未來之後續探討可參考此研究之巢狀網格、數值潮位計選擇、簡化地形之設置等，透過分析變量之於波速之無因次參數I_z之關係，得知各種變量對於水位之時差影響，以期於未來能更準確地分析水位，以及將經驗公式之設計方法應用於多個領域當中，包含河口地區與河川之相關評估，或潮汐發電之波速、潮汐到時評估等。;Previous studies have pointed out that discrepancies may arise between the total water level and the observed data after accounting for residuals when examining storm surge water levels in model simulations. Possible factors contributing to these discrepancies include the impact of extreme events, unique topographical features of specific areas, and the positioning of tide gauges inside harbors rather than in open waters. Therefore, it is necessary to examine the effects of pure storm surge water and astronomical tidal levels separately. However, even when examining tidal water levels alone, fluctuations in water level values can still occur, indicating that factors other than storm surges may affect the accurate identification of water levels.
To understand the water level residuals, this study employed the COMCOT-SS numerical model (Cornell Multi-grid Coupled of Tsunami Model—Storm Surge) along with the TPXO8-atlas astronomical tide model as boundary conditions. By solving nonlinear shallow water equations, the study simulated water level results at several existing tide gauge stations during specific pure tidal periods and conducted convergence analysis of statistical parameters, revealing the potential impact of tidal arrival times on total water level residuals.
Various variables were considered in exploring the temporal effects of harbor structures on water levels, including Manning′s coefficient, wave height, the distance between tide gauges, and water depth. To establish relationships between these parameters, this study conducted a dimensionless analysis of wave speed using the simplified characteristics of harbor topography. The dimensionless parameter I_z (calculated as I_z=C_s/C_p, where C_s is the simulated wave speed obtained from the distance between two different tide gauges and the arrival time of the incoming wave; C_p is the predicted wave speed calculated using the formula C_p=√gH, where g is the gravitational acceleration, H is the water depth) was used to record the effects of harbor structures on the time differences in water levels inside and outside the harbor, observing the potential impact of harbor structure shielding on wave speed simulations. Subsequent water level analyses were conducted for actual harbor cases to assess the influence of different variables on the timing of water levels.
The study found that the dimensionless parameters x_1, x_2 and x_3, derived from three different variables (wave height h, Manning′s coefficient n, and water depth H), could potentially contribute to timing errors in water levels. It was also noted that the dimensionless parameter x_2, calculated from Manning′s coefficient n, was negatively correlated with the dimensionless wave speed parameter I_z. In contrast, the dimensionless parameter x_1, calculated from wave height h, was positively correlated. Ultimately, an empirical formula was derived through linear regression analysis of all data points, which, when compared with actual harbor simulation cases, confirmed its applicability in assessing the differences in wave speeds and timing between tide gauges inside and outside harbors.
Future research can reference this study′s use of nested grids, selection of numerical tide gauges, and simplified terrain settings. Analyzing the relationship between variables and the dimensionless parameter I_z of wave speed can achieve more accurate water level analyses. Furthermore, the method of designing empirical formulas developed in this study could be applied to various fields, including estuarine and river assessments or evaluations of wave speed and tidal arrival time for tidal energy generation. |
