氣象合成與水文模式之發展及因應氣候變遷之供水系統調適能力建構

Abstract

臺灣地區地形特殊導致80%左右之逕流量直接入海，加上豐水期及枯水期雨量比例差異大，使得水資源之調配相當不易，氣候變遷可能導致豐枯水期降雨量改變，造成豐枯水期流量差異擴大，更加劇了供給足夠水量的困難度。為了評估氣候變遷對供水系統造成的影響，並提出調適措施以降低氣候變遷可能帶來的衝擊，必須發展氣候變遷對供水系統衝擊評估與調適能力建構之流程與工具。本研究在流程上依據聯合國發展規劃署所提出的「氣候變遷調政策略綱領」分成五大步驟，由問題界定開始，系統性的評估現況與未來的脆弱度與風險，接著提出解決問題之可能之調適策略及其評估方式，最後並持續調適措施之落實與修正。評估工具方面，本研究注重於發展可同時繁衍產生日間尺度及小時時間尺度之氣象合成模式與可同時模擬日時間尺度及小時時間尺度之水文模式，配合研究區域之水資源供水系統模式，可分析氣候變遷對供水系統之衝擊。 本研究中建立雙機率分佈之氣象合成模式，將降雨事件分為暴雨事件及一般降雨事件，利用暴雨事件發生機率決定降雨日之降雨量是否為暴雨事件，並採用不同的機率分佈描述暴雨及非暴雨事件之降雨，模式中利用Huff法建立雨型，當發生暴雨事件時可配合雨型之挑選決定24小時之雨量。此外，研究中建立以Green-Ampt法計算入滲量與結合水平衡運算之水文模式，該模式可計算以日為時間尺度之河川流量，亦可配合雙機率分佈之氣象合成模式提供之含有日及小時時間尺度之氣象資料，在非暴雨事件時以日為時間尺度模擬河川流量，而在暴雨事件發生時則以小時時間尺度模擬小時河川流量。配合不同時間尺度之流量資料，建立同時考慮日及小時時間尺度之水庫操作模式，在暴雨事件發生時以防洪操作進行水庫放水，而非暴雨時期為一般放水操作之水庫操作模式。 根據分析顯示，氣候變遷將使得石門水庫上游河川流量在豐水期呈現增加趨勢而枯水期呈現減少趨勢，且造成未來水庫水位之50百分位之水庫水位低於下限的旬數情況將較歷史情況為多，而各旬最大水位分析結果指出在豐水期時水庫最大水位可能超過水庫大壩安全容許之最高水位，顯示未來水庫不論是在水資源調配或是防洪操作上將面臨更加嚴格的挑戰。而氣候變遷對臺北供水區衝擊影響不大，可維持正常供水且有多餘水量可提供其他地區使用，板新供水區雖可滿足在容忍度SI為0.5之供水情形，但偶爾會有缺水事件發生，而在桃園供水區在容忍度SI為0.5的條件下，供需缺口最高達44萬噸/日，且由回復力指標可發現桃園供水區長期處在缺水的情況，加入既有策略後可提供之水量可滿足桃園供水區之需水量，但仍可能面臨短期缺水事件之風險，因此可採取埤塘、水庫清淤、淨水場供水能力增加與汰換舊漏自來水管線等慮強化方案以降低未來面臨之風險。 本研究所建立之氣象合成模式可適切的繁衍產生氣候變遷情境下之氣象資料，透過本研究建立之水文模式可模擬未來河川流量，再利用水庫操作模式可評估氣候變遷對水庫功能與供水系統之衝擊。而利用本研究之調適策略評估流程，則可有系統的分析及評估可能的衝擊與挑選可能之調適策略，以降低氣候變遷可能造成的衝擊。

Because of the 80% of stream flow runs into the ocean directly and the difference of rainfall among wet and dry seasons is significant, the allocation of water resource is very challenge. In addition, the climate change may lead to the changes of rainfall and streamflow and enlarge the difference amnong wet and dry seasons. These may exacerbate the difficulty of supplying sufficient water. To reduce the impacts of climate change on the water supply system and to develop effective adaptation measures, this study follows Adaptation Policy Frameworks for Climate Change (APF) proposed by United Nations Development Programme (UNDP) and focuses on the development of novel weather generator and hydrological model. The developed weather generator can provide daily series with embeded hourly rainfall data for storm events. The proposed hydrological model can run in daily and hourly time steps depending on its input weather data. The simulated streamflows can be used as inputs for reservoir operation model and water resource system dynamics model to assess the climate change impacts. The novel weather generator built in this study uses dual probability distributions to describe the extreme rainfall and normal rainfall events respectively. The probability of extremely rainfall event for each month is further used to decide if the rainy day is a strom event or not. Besides, the hyetographs are described by the Huff method and the 24 hourly rainfalls are generated by a selected hyetograph. On the other hand, the Green-Ampt method and water balance equations are used to develop NTU_Wtaershed Hydrological Model (NTU_WH). Not only the daily stremaflow but also the hourly dtreamflow in the extremely rainfall event days can be calculated by the NTU_WH. The reservoir operation model is also developed in this study, which can be run with different time scales. This reservoir operation model release houly flood water in the extreme rainfall event days and daily water supply in non-extreme rainfall event days. To assess the impact and the adaptive capacity of climate change on water supply system, the adaptation strategies assessment process follows APF. APF has five major steps, including (1) defining project scope and design, (2) assessing vulnerability under current climate, (3) characterizing future climate related risks, (4) developing an adaptation strategy, and (5) continuing the adaptation process. According to the simulations of streamflow and water usage under different climate change secnarios, the vulnerability of water supply systems for different districts can be estimated. Furthermore, the apative stretageies can be evaluated to reduce the deficits of water supply and the vulnerability for different water usages. According to the results, the inflow of Shihmen Reservoir may increase in wet season and decrease in dry season under different future climate change scenarios. These results lead to the lower reservoir storages in dry season and the maximal reservoir storage for each 10-days in the wet season may be higher than the permitted maximal storage of dam safety. The water resource reallocation or the flood mitigration may become severe in future. However, the impact of climate change on carrying capacity of Taipei city is not significant. The carrying capacity of water supply with the criterion of Shortage Index (SI) =0.5 for Banshin area is enough to meet water demands, but the events of water shortagecan still be expected. In Taoyuan water District, under the criterion of SI=0.5, the maximal deficits of water supply and demand is 440,000 tons / day. Besides, the Taoyuan area is always in the condition of insufficient water supply according to the index of resilience. Fortunately, if the planned water strategies for the Taoyuan area can be implemented, the gap between water supply and demand can be bridged. Although the gap could be filled, the risk of short-term water shortage in the Taoyuan area is still higher. The strengthening strategies could be adopted in this area are ponds, reservoirs desilting, increasing the capacity of water purification plant, and the replacement old water pipeline. The weather generator built in this study can generate the weather series for different climate change scenarios. The streamflows can be simulated by the NTU_WH which is also developed in this study. Then, the impacts of climate change on reservoir functions can be evulated by the reservoir operation model. Besides, the adaptation strategy assessment process used in this study can assess the impacts of water supply system on climate change and evulate the possible adaptation strategies to reduce the potential impacts of climate change.