結合空間–時間統計降尺度模式和降雨逕流模式評估時逕流

Integration of spatio-temporal statistical downscaling and rainfall-runoff modeling for estimating hourly streamflow

陳語庭(Yu-Ting Chen)

指導教授 ： 林國峰

國立臺灣大學/工學院/土木工程學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201703556

摘要

大氣環流模式配合降尺度獲得的高解析度雨量和溫度資料是評估氣候變遷對降雨逕流衝擊的重要一環，然而現存的溫度降尺度模式未考慮跨日關係與日夜循環，且降雨逕流模式鮮少評估小時尺度。因此，本研究提出同時考慮跨日關係及日夜循環的統計降尺度模式和降雨逕流模式，評估氣候變遷對小時逕流的衝擊。
首先本研究提出溫度降尺度模式可分為空間和時間降尺度階段。空間降尺度階段，使用K最近鄰居法(k-nearest neighbor method, KNN)建構大尺度大氣因子與溫度的關係，獲得測站尺度日溫度序列。時間降尺度階段使用KNN結合基因演算法(genetic algorithm, GA)，同時考慮跨日關係及日夜循環，由日溫度降尺度至小時溫度。進而將繁衍的溫度和雨量輸入HBV模式(Hydrologiska Byråns Vattenbalansavdelning)模擬日和小時的逕流量，最後評估氣候變遷對逕流的衝擊。本研究將提出之模式應用於臺灣石門水庫流域，模式建立使用大尺度大氣因子為NCEP/NCAR (NCEP/NCAR reanalysis data)再分析資料，及測站尺度日與時觀測資料。未來情境映射使用BCM2.0模式中的A2、A1B及B1情境之中期(2046-2065)和長期(2081-2100)資料。
結果顯示，本研究提出之統計降尺度模式能有效保有日和小時溫度的統計特性，並保持跨日關係與日夜循環，HBV模式可準確模擬小時的逕流。未來情境映射時，所有情境的溫度會增加，逕流會減少。本研究對溫度和降雨逕流評估結果，可幫助決策者進行水資源系統規劃及管理。

關鍵字

氣候變遷 ； 溫度 ； 逕流 ； 統計降尺度 ； K最近鄰居法 ； 基因演算法 ； HBV模式

並列摘要

Downscaling is used to translate the general circulation model outputs to a finer spatiotemporal resolution rainfall and temperature data, which is essential for assessing hydrological impacts of climate change. However, existing downscaling models for hourly temperature are unable to consider the inter-daily connection and the diurnal cycle and consider the hourly scale. Besides, existing rainfall-runoff models have less attention on hourly streamflow. To address this problem, a spatiotemporal downscaling model is proposed which is capable of reproducing the inter-daily connection, the diurnal cycle, and the statistics on daily and hourly scales. Using the results of rainfall and temperature as input to the rainfall-runoff model, the hydrological impacts of climate change on hourly streamflow are assessd.
The proposed downscaling model consists of two steps, the spatial downscaling and temporal downscaling. The spatial downscaling is applied first to obtain the relationship between large-scale weather factors and daily temperature at station scale using the k-nearest neighbor method. Then, the hourly downscaling of daily temperature is conducted in the second step using the k-nearest neighbor method with the genetic algorithm and consideration of the inter-daily connection and the diurnal cycle. The large-scale datasets, which are obtained from the National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis data and the Bergen Climate Model version 2 (BCM2.0) outputs, and the local temperature data are analyzed to assess the impacts of climate change on temperature. After rainfall and temperature projections, the HBV model are applied to predict the future streamflow of Shihmen reservoir basin (Taiwan) under the A2, A1B and B1 scenarios of the BCM2.0 for the periods 2046-2065 and 2081-2100.
The results show that the proposed model can accurately reproduce the local temperature and its statistics on daily and hourly scales. The change trend of the future streamflow under the impacts of climate change is presented and discussed. To conclude, the results demonstrate that the proposed model is a powerful tool for predicting streamflow. The future changes of temperature and streamflow statistics are obtained which will help decision makers in planning and management of water resources systems.

並列關鍵字

Climate change ； Temperature ； Streamflow ； k-nearest neighbor method ； HBV model

參考文獻

81. 吳爵廷，2014，基於雨量型態的統計降尺度方法研究，國立台灣大學土木工程學系碩士論文。

1. Arda Şorman, A., Şensoy, A., Tekeli, A.E., Şorman, A.Ü. and Akyürek, Z., 2009. Modelling and forecasting snowmelt runoff process using the HBV model in the eastern part of Turkey. Hydrological Processes. 23, 1031–1040.

2. Anandhi,A., Srinivas, V.V., Nagesh Kumar, D. and Ravi, S.N., 2009. Role of predictors in downscaling surface temperature to river basin in India for IPCC SRES scenarios using support vector machine. International Journal of Climatology. 29, 583–603.

4. Hewitson, B.C. and Crane, R.G., 2005. Consensus between gcm climate change projections with empirical downscaling: precipitation downscaling over south africa. International Journal of Climatology. 26, 1315–1337.

5. Belli, A. and Haberlandt, U., 2012. Stochastic precipitation modeling using circulation patterns to analyze climate impact on floods. Adv. Geosci. 32, 93–97.