Nearshore Wave Height Hindcasting by Using a Coupled Numerical-Statistical Prediction Model during Typhoons

View Statistics Email Alert RSS Feed

Information

Details

Project title

Code/計畫編號

MOST105-2221-E019-041

Translated Name/計畫中文名

耦合數值統計模式預測近岸颱風風浪之研究

Project Coordinator/計畫主持人

Chih-Chiang Wei

Funding Organization/主管機關

National Science and Technology Council

Co-Investigator(s)/共同執行人

張人懿

Department/Unit

Department of Marine Environmental Informatics

Website

https://www.grb.gov.tw/search/planDetail?id=11906513

Year

2016

Start date/計畫起

01-08-2016

Expected Completion/計畫迄

31-07-2017

Bugetid/研究經費

800千元

ResearchField/研究領域

土木水利工程

"本計畫擬建立一耦合數值統計模式以便在颱風時期即時預測近岸波浪。此一概念乃混合了數值模式與統計模式並結合兩者之間的特性，可用以解決即時預測近岸風浪問題。一般而言，數值模式優點為所有參數都是由動力方程式計算得出的，其預報結果可由物理關係解釋，因此數值模式可用以求解高度非線性的波浪行為。在數值模式中，觀測資料可提供動力模式合理的初始條件，在其演算過程中觀測資料則使用上較少。然而，數值模式模擬過程常須耗費大量計算時間，在模擬短期距（如15 分鐘）波浪預測時，可能無法提供即時性；相對的，統計模式在計算效率則較數值模式為高，其因是統計模式乃由過去長期觀測資料建立預測變數與其他屬性間之統計關係，利用合適的統計方法可快速地建立預測模式。然而，對數值模式而言，統計模式之物理基礎較為薄弱，如兩參數統計關係很強但其物理上可能無因果關係可解釋，或是物理上顯而易見的兩參數關係亦可能未反應在統計分析上，因此統計模式有其侷限性。本計畫擬以臺灣東北部近岸海域為研究例。本計畫發展耦合數值統計模式過程中，首先利用 SWAN（Simulating WAve Nearshore numerical model）數值模式模擬臺灣東北部周圍海域的海象波高、波向和週期。SWAN 數值模式為Booij et al. (1996)所發展，適用於海岸地區、湖泊或河口附近水域之波浪模擬，屬於第三代的風浪預報模式。SWAN 模式考慮波浪在時間及空間領域中的傳遞、波與波的非線性交互作用、波浪受風的成長、碎波、因底床摩擦所造成的能量衰減、及受到海流及地形變化而產生的頻率位移、淺化與折射、波浪所引致的平均海水面的上升等，因此計畫中SWAN 模式將用來推算近岸波浪，SWAN 數值模式模擬時間將需數小時。本文將收集2000 至2015 共41 場颶風資料以及5 個浮標海象資料（富貴角、基隆、龍洞、龜山島及蘇澳等測站），並設計4 個近岸任意點（樣本點）以為試驗點（無測站海象資訊）。首先，本計畫利用SWAN 數值模式模擬多場颱風網格點波高值，並得到所有浮標點和樣本點之波高數值解值；其次，本計畫以新近類神經網路模式的調適性模擬推論系統網路(Adaptive Network-based Fuzzy Inference System, ANFIS) 建構近岸風浪統計預測模式，ANFIS 網路為結合類神經網路優點（如學習能力、最佳化能力、連結式的結構）以及模糊邏輯優點（如接近人類的思考行為，容易結合專家知識），因此近年來被廣泛地應用在各個領域上。整合模式之輸入資料為颱風資料以及浮標海象資料；最後，本計畫將比較浮標點的波高觀測值、數值解以及統計解，以驗證統計解的合理性。任意樣本點則因缺乏觀測資料，因此波高值將藉由高精度數值解（視為真值）替代，因此樣本點所建構的預測模式輸入資料將為颱風資料和數值模式所產生的海象參數。本計畫預期將完成上述預測模式驗證，同時討論各模式預測能力的優缺點及適用性，以提供臺灣近岸海域船舶與人員航行安全。""The challenges of achieving wave height predictions when shipping affected by typhoons. A numerical model coupled with a data-driven statistical method is proposed for forecasting the wave climate of a shipping route during hurricanes. The developed model can be used to determine the wave heights on a ship’s trajectory, considering a short time step (15 min) of a ship’s operation. We used an artificial neural network (ANN) based Adaptive Network-based Fuzzy Inference System (ANFIS) to build a statistical prediction model. The SWAN (Simulating WAve Nearshore numerical model) developed by Booij et al. (1996) will be employed as the numerical simulation based model. The proposed coupled numerical-statistical prediction model combines the NUM-based with the ANN-based model, and can be used to determine the wave heights of shipping points where buoy measures are absent. The experimental area near Northeast Coast of Taiwan will be used for simulation. Regarding a complex typhoon system, the collected data comprise the typhoon characteristics, the buoy atmospheric properties, and the buoy maritime properties. The 5 buoys, namely, Fugui Cape, Keelung, Longdong, Guishandao, and Su-ao, will be selected for data collection. This study includes 41 typhoon events where the typhoon tracks classified as Types 14 and 6 over the past 16 years (2000–2015). Type 1 is the direction of westward movement traveling through the southern East China Sea (near Northern Taiwan), Types 24 are the direct westward movement across the Central Mountain Range (CMR) of Taiwan, and Type 6 is the northward movement along the eastern coast of Taiwan. Four criteria, including the relative mean absolute error, coefficient of variation of the root mean squared error, correlation coefficient, and efficiency coefficient, will be used to highlight the scenario capable of identifying the optimal performance level. This study will compare the SWAN numerical model, the ANFIS statistical model, and the coupled SWAN and ANFIS model with observations. This study will identify the optimal model cases. Several typhoons will be used as simulation typhoons for the real-time wave height forecasts. This study will demonstrate the feasibility of the proposed methodology and discuss the scenarios on the basis of increasing the accuracy and efficiency of predictions."

DSpace CRIS

Nearshore Wave Height Hindcasting by Using a Coupled Numerical-Statistical Prediction Model during Typhoons

Details

Description