應用空間系統動力於評估水文服務及其不確定性分析

Abstract

水是維持人類生命不可或缺的元素，同時也是生態系所提供的重要服務之一。但大量的人為開發，如坡地濫墾、森林砍伐、過度施肥等已超越生態系原有之承載力，導致受損的生態系無法正常發揮調節洪峰、水土保持、過濾水質等功能，對人類生活造成衝擊，造成惡性循環。 水文水理與水文服務猶如孿生兄弟，前者以定量方式描述流量、水質等時序上之變化；後者則偏向以定性方式呈現其空間分布特性。隨著生態水文學之興起，Hydro-ecology與Eco-hydrology這類詞彙開始出現於相關研究內。Zalewski與林裕彬等人將生態水文學明確定義為「一門在不同時空間尺度下，結合水文過程及生物動態變化的跨領域科學」。本研究以改良後同時具有時空間特性之空間系統動力(SSD)模擬平台，串聯起水文水理模式(GWLF)及水文服務評估工具(InVEST)，作為研究生態水文之媒介。 選定南投縣陳有蘭溪作為研究區域，以實測資料進行流量驗證，本研究建立之HSSSD其RSQ與NSE都逼近0.85，相較於InVEST的RSQ=0.69/ NSE=0.65與GWLF的RSQ=0.79/ NSE=0.78，顯示HSSSD能有效提升模擬之準確性。再者，本研究以GLUE方法進行參數之不確定性分析，顯示對應土地利用為森林之CN2、r(退水係數)，以及D(深層滲漏)都對模擬之結果有較大之不確定性。 結合水文水理模式與水文服務評估工具，HSSSD可同時展現時空間尺度之模擬結果，亦可提供使用者如營養鹽流失途徑等更多資訊，除進一步提升模擬之準確性，更以不確定性分析劃設模擬之95%信賴區間，使模擬結果能被零活應用於生態水文學之中。

Water is an indispensable element in the maintenance of human life and one of the important services provided by the ecosystem. However, a great number of artificial development, such as sloping land, deforestation, excessive fertilization, etc., have surpassed the original carrying capacity of the ecosystem, resulting in the failure of the ecosystem to function properly to regulate flood peaks, soil and water conservation, and to filter water quality. Causes an impact and creates a vicious circle to human beings. Hydrologic and Hydraulic Modeling and Hydrology Services are like twin brothers. The former describes the changes in flow rate and water quality in a quantitative way; the latter tends to present its spatial distribution characteristics in a qualitative way. With the rise of eco-hydrology, terms such as “Hydro-ecology” and “Eco-hydrology” began to appear in related research. Zalewski, Lin Yu-Pin and others have clearly defined ecohydrology as 'A cross-disciplinary science that combines hydrological processes and biological dynamics at different spatial scales.' In this study, the spatial system dynamics (SSD) simulation platform with improved time-space characteristics is combined with hydrological and hydrological model (GWLF) and hydrological service assessment tool (InVEST) as the medium for studying ecological hydrology. Chen You-lan river, Nantou County was selected as the research area, and the simulation result was validated by the measured data. The HSSSD , which was built in this study, its RSQ and NSE approaching 0.85, compared with RSQ=0.69/ NSE=0.65 of InVEST and RSQ=0.79/NSE=0.78 of GWLF., showing that HSSSD can effectively improve the accuracy of the simulation. Furthermore, this study uses the GLUE method to analyze the uncertainty of the parameters, showing that the corresponding is CN2, r of the forest and D. Combined hydrological and hydraulic models with hydrological service assessment tools, HSSSD can simultaneously display simulation results at time and space scales, and provide users with more information such as nutrient loss pathways, in addition to further improving the accuracy of simulations and uncertainty. Analyze the 95% confidence interval of the simulation, so that the simulation results can be flexibly applied to ecological hydrology.