結合水質與粒子追蹤模擬探討颱風事件後感潮河段之魚群死亡機制

Abstract

颱風及強降雨事件期間，有機物與污染物往往因降水沖刷而從上游及周邊環境大量輸入河川系統。在感潮河段，受漲退潮機制影響，污染物可能滯留並延遲輸移過程，進而對水質條件及河川生態系統造成嚴重衝擊。國內外研究案例顯示，暴風雨事件過後，河川經常發生大規模魚類死亡的現象。位於淡水河系中游且經密集都市區域的基隆河下游，即於2015年7月昌鴻颱風事件不久後發生魚群死亡事件。淡水河為臺灣北部重要的河口生態系統，然而，其中下游都市化河段長期面臨都市廢水、工業排放、畜牧廢水及非點源污染等多重壓力，水質問題顯著。尤其在颱風過後，河川流量迅速減少，大量污染物滯留感潮區段，導致溶氧（Dissolved Oxygen, DO）迅速耗盡、污染濃度超出河川自淨能力，進一步造成水質急遽惡化。為探討颱風事件期間及其過後河川水質變化過程，以及導致魚類大規模死亡的成因，本研究採用EFDC數值模式進行水動力與水質模擬，並結合拉格朗日粒子追蹤模型模擬魚群死亡後的漂移軌跡。 研究以2015年基隆河烏魚大規模死亡事件為案例，整合淡水河系於昌鴻與蘇迪勒颱風期間的水理與水質數據，建置並驗證水動力模型（HDM）與水質模型（WQM）。為率定與驗證HDM及WQM參數，研究透過多項統計指標分別對水位、氨氮（Ammonia Nitrogen, NH3-N ）和溶氧的模擬結果進行誤差分析。結果顯示，所建立之HDM與WQM具有良好的可靠性與可信度（HDM的R-squared均大於0.9，WQM的MAE均小於0.6 mg/L），可有效模擬颱風事件期間的水理與水質變化。魚群死亡事件之水質模擬結果顯示，颱風期間高流量攜帶了高負荷的營養鹽和有毒物質進入河川中，因流量大導致稀釋效應明顯，污染物濃度相對較低，水質在短期內維持良好狀態。然而，颱風過後流量驟減，受潮汐作用影響水動力大幅減弱，導致河段內有機物與污染物的累積與滯留現象加劇，加速氧氣消耗，而氨氮濃度則顯著上升，形成對魚類不利的複合致死條件。經研究推算，魚類之48小時非離子態氨（NH3）半致死濃度（LC50）為0.092 mg/L，約為水溫攝氏30度且pH值為7.4的總氨氮濃度4.551 mg/L中所含之NH3，因此以NH3-N濃度4.551 mg/L作為氨氮警戒值；溶氧則以魚類缺氧窒息濃度0.9 mg/L為溶氧警戒值。以昌鴻颱風為例，模擬結果顯示基隆河中NH3-N濃度高於魚類48小時LC50時長超過60小時，DO濃度低於0.9 mg/L的持續時間亦達相同長度，驗證水質劣化對魚類生存的威脅。此外，粒子追蹤模型詳細模擬魚群死後的移動軌跡，並根據模擬結果推測魚群死亡地點位於大直橋附近，且烏魚浮起後首次被發現的時間和位置分佈均與事證紀錄大致相同，顯示模擬具備良好的準確性與應用價值。進一步分析指出，若將昌鴻颱風過後至魚群死亡事件期間，基隆河低於20 cms之入流量均調升為20 cms之定流量，DO濃度可由長時間低於1 mg/L顯著提高為1.5∼2.5 mg/L，而NH3-N濃度則由高於4.551 mg/L降低至3∼4 mg/L，有效提升河川涵容能力（assimilation capacity）並降低魚類死亡風險。本研究結果表明，颱風過後適當的流量調控在維護河川生態穩定中具有關鍵作用，為河川管理與生態保護提供重要參考依據。

During typhoons and intense rainfall events, substantial amounts of organic matter and pollutants are often introduced into river systems from upstream and surrounding areas via surface runoff. In tidal-affected reaches, tidal fluctuations can lead to pollutant retention and delayed transport processes, thereby exerting severe impacts on water quality and river ecosystems. Both domestic and international studies have documented frequent occurrences of mass fish kills following storm events. For example, in July 2015, shortly after Typhoon Chanhom, a significant fish kill was observed in the downstream Keelung River, which traverses highly urbanized areas in the midstream of the Tamsui River system. The Tamsui River, a critical estuarine ecosystem in northern Taiwan, faces persistent stressors, including municipal wastewater, industrial effluents, livestock waste, and non-point source pollution in its urbanized downstream segments. These pressures have resulted in prominent water quality challenges. Following typhoons, the rapid decrease in river discharge often results in pollutant retention within tidal reaches, leading to severe oxygen depletion and pollutant concentrations that surpass the river’s self-purification capacity, causing abrupt water quality deterioration. This study examines the processes of water quality variations during and after typhoon events, along with the mechanisms underlying large-scale fish kill events. The Environmental Fluid Dynamics Code (EFDC) numerical model was utilized to simulate hydrodynamic and water quality dynamics. Furthermore, a Lagrangian particle tracking model was integrated to simulate the drift trajectories of fish carcasses following mortality events. The results revealed that high discharge during typhoons transported significant loads of nutrients and toxic pollutants into the river. However, due to substantial dilution effects from elevated flow, pollutant concentrations remained relatively low, temporarily maintaining good water quality. In contrast, post-typhoon conditions saw a rapid reduction in discharge, coupled with weakened hydrodynamic activity due to tidal influence. This led to the accumulation and retention of organic matter and pollutants, accelerating oxygen depletion and significantly increasing ammonia nitrogen levels, creating lethal conditions for fish. For instance, during Typhoon Chanhom, simulations indicated that ammonia nitrogen concentrations in the Keelung River exceeded the 48-hour LC50 (median lethal concentration) for fish for over 60 hours. Similarly, dissolved oxygen levels dropped below 0.9 mg/L for a comparable duration, confirming the threats posed by water quality degradation to fish survival. The particle tracking model accurately simulated postmortem drift trajectories, identifying the likely location of fish kills near Dazhi Bridge.The timing and spatial distribution of floating fish matched observational records, demonstrating the model’s robustness and applicability. Further analysis suggested that increasing the Keelung River’s inflow following Typhoon Chanhom could significantly elevate dissolved oxygen levels, reduce ammonia nitrogen concentrations, and enhance the river’s assimilative capacity, effectively mitigating fish mortality risks. During Typhoon Soudelor, sustained higher inflows significantly improved post-typhoon water quality conditions and reduced mortality incidents. These findings underscore the critical role of flow regulation in maintaining ecological stability in rivers following typhoon events. This study provides valuable insights for river management and ecological conservation, emphasizing the importance of adaptive flow control strategies.