目前氣候變遷情境下的衝擊研究大多根據大氣環流模式(general circulation model, GCM)的模擬結果進行研究，唯因GCM屬於全球尺度的模擬與推估，其模擬值可能無法合理描述小尺度區域的變遷特性，故可透過降尺度(downscaling)過程推求能代表小尺度區域特性的氣候變遷資訊。本文發展以日為時間單位的降尺度模式，包含多變量線性降尺度法與支撐向量機降尺度法，並與國際常用之統計降尺度套裝軟體SDSM (Statistical Downscaling Model)進行比較，主要討論石門水庫集水區的雨量變遷情況。SDSM利用敏感度分析挑選適當預測因子，再以序率合成及轉換函數法進行降尺度分析。本研究發展的兩模式基於統計理論建立降尺度模式並依據氣象條件將某日分類判斷為降雨日或不降雨日，若當日為降雨日，再以回歸模式推估降雨日的降雨量。
在模式驗證結果以各月平均雨量顯示，支撐向量機降尺度法有較好之模擬結果，而對於未來模擬情況則三種方法趨勢不盡相同，目前GCMs對於未來氣候變化情境下降雨預測存在高度不確定性，此三種降尺度模式是進行降雨時空模擬，分析小尺度下未來降雨變化可能情境的一種較為簡單、經濟的方法，可作為以後評估水資源衝擊之參考。

The general circulation models (GCMs) outputs are currently recognized as convinced projected data for climate change studies. However, GCMs are global-scale models and may not properly describe climate change characteristics of local regions, such as catchments in Taiwan. Therefore, downscaling methods were proposed to downscale large-scale GCMs outputs into information of small-scale regions that can represent climate change characteristics of local catchments. The study developed a multiple regression downscaling model and a support vector machine (SVM) downscaling model, and compared those developed models with a popularly used software, Statistical Downscaling Model (SDSM). SDSM uses sensitive analysis to choose predictors, and applies stochastic generation and transform function to produce downscaled data. The developed multiple regression and SVM downscaling models are both based on statistical methods, with statistical tests to select significant predictors and a multiple regression function to downscale GCMs outputs into small-scale catchment data. The three methods are applied to rainfall downscaling in Shihmen reservoir catchment. Analysis results show that projected rainfalls from three methods outperform one another in different rainfall characteristics. Among them, SVM downscaling model is better than other methods in simulating average monthly rainfall. Downscaling GCM projected rainfalls for hydrological uses still has uncertainties and needs further investigations in the field of water resources. This study developed two statistical downscaling models that can contribute to researches of climate change impacts on water resources.

參考文獻
方世榮，1933，「統計學導論」，華泰書局。
林傑斌、陳湘、劉明德，2002，「SPSS11統計分析實務設計寶典」，碩博文化。
張逸凡，2005，「支撐向量機在即時河川水位預報之應用」，碩士論文，國立成
功大學水利及海洋工程研究所。
陳憲宗、游保杉，2007，「洪水位之即時機率預報 - 結合支撐向量機與模糊推
理」，農業工程學報，第53卷，第4期，第1–20頁。
童慶斌、游保杉、李明旭，2007，「強化區域水資源永續利用與因應氣候變遷之調適能力(1/2)」，經濟部水利署水利規劃試驗所出版。
顏月珠，1986，「實用無母數統計方法」，陳昭明發行，台北市。
Aavudai Anandhi, V. V. Srinivas, Ravi S. Nanjundiah and D. Nagesh Kumar (2007). “Downscaling precipitation to river basin in India for IPCC SRES scenarios using support vector machine.” International Journal of Climatology (in press), 28, 401–420.
Cannas, B., A. Montisci, A. Fanni, L. See, and G. M. Sechi (2004). “Comparing Artificial Neural Networks and Support Vector Machines for Modelling Rainfall-Runoff.” In: Proceedings of the 6th International Conference on Hydroinformatics, S. Y. Liong, K. K. Phoon, and V. Babovic, eds., (Singapore, 21–24 June 2004), World Scientific Publishing Co., Singapore.
Cavazos T and Hewitson BC. (2005). “Performance of NCEP-NCAR reanalysis variables in statistical downscaling of daily precipitation.” Climate Research,28, 95–107.
Chang, C. C. and C. J. Lin (2002). “Training ν -Support Vector Regression: Theory and Algorithms.” Neural Computation,14(8), 1959–1977.
Cherkassky, V. and Y. Ma (2004). “Practical selection of SVM parameters and noise estimation for SVM regression.” Neural Networks,17, 113–126.
Fisher,R.A. (1936). “The use of multiple measurements in taxonomic problems.” Annals of Eugenics,7, 179–188.
Fletcher (1987) Practical Methods of Optimization, John Wiley and Sons, New York.
Fowler HJ, Ekström M, Kilsby CG, and Jones PD. (2005). “New estimates of future changes in extreme rainfall across the UK using regional climate model integrations.1.Assessment of control climate.” Journal of Hydrology,300, 212–233.
Frei C, Christensen JH, Déqué M, Jacob D, Jones RG, and Vidale PL. (2003). “Daily precipitation statistics in regional climate models:evaluation and intercomparison for the European Alps.” Journal of Geophysical Research, 108(D3):4124, DOI: 10.1029/2002JD002287.
Frei C, Schöll R, Fukutome S, Schmidli J, and Vidale PL.(2006). “Future change of precipitation extremes in Europe: An intercomparison of scenarios from regional climate models.”Journal of Geophysical Research-Atmospheres,111: D06105, DOI: 10.1029/2005JD005965.
G. Bürger (2005).”Regression-based downscaling of spatial variability for hydrologic applications. “Journal of Hydrology,311, 299–317.
Han, D. and Cluckie, I. (2004). “Support vector machines identification for runoff modeling. “In: Proceedings of the 6th International Conference on Hydroinformatics, S.Y. Liong, K.K. Phoon and V. Babovic, eds., (Singapore, 21–24 June 2004), World Scientific Publishing Co., Singapore.
Hsu, C. W., C. C. Chang, and C. J. Lin,(2003).”A Practical Guide to Support Vector Classification.”Technical report, Department of Computer Science and Information Engineering, National Taiwan University, Available at: http://www.csie.ntu.edu.tw/~cjlin/papers/guide/guide.pdf
Huth R. (1999).”Statistical downscaling in central Europe: evaluation of methods and potential predictors. “Climate Research,13, 91–101.
IPCC (2007). “Climate Change 2007：Climate Change Impacts, Adaptation and Vulnerability.”, Working Group II Contribution to the Intergovernmental Panel on Climate Change Fourth Assessment Report.
Lin, C. J. and co-workers (2004). LIBSVM Software (Version 2.71, released on November 20, 2004), Available at:
http://www.csie.ntu.edu.tw/~cjlin/libsvm/index.html
Liong, S. Y., and C. Sivapragasam (2002).”Flood Stage Forecasting with Support Vector Machines. “Journal of the American Water Resources Association,38(1), 173–186.
Prudhomme C, Reynard N, and Crooks S. (2002). “Downscaling of global climate
models for flood frequency analysis: Where are we now?. “Hydrological Processes,16, 1137–1150.
Radan Huth (2005). “Downscaling of humidity variables: A search for suitable predictors and predictands . “International Journal of Climatology ,25, 243–250.
Robert L. Wilby, and Christian W. Dawson (2007). “SDSM 4.2 — A decision support tool for the assessment of regional climate change impacts .”,User Manual, 1–94.
Salathé EP. (2005). “Downscaling simulations of future global climate with application to hydrologic modelling.”International Journal of Climatology,25, 419–436.
Subimal Ghosh, and P.P. Mujumdar (2008).”Statistical downscaling of GCM simulations to streamflow using relevance vector machine. “Advances in Water Resources,31, 132–146.
Tripathi, S., Srinivas, V.V., Nanjundiah, and R.S. (2006). “Downscaling of precipitation for climate change scenarios: A support vector machine approach”Journal of Hydrology,330(3), 621–640.
Vojinovic, Z., and V. Kecman (2004).”Contaminant Transport Modelling with Support Vector Machine Model - An Alternative to Classical Advection-Dispersion Equation.”In: Proceedings of the 6th International Conference on Hydroinformatics, S. Y. Liong, K. K. Phoon, and V. Babovic, eds., (Singapore, 21–24 June 2004), World Scientific Publishing Co., Singapore.
Widmann M., and Bretherton CS. (2000).”Validation of mesoscale precipitation in the NCEP Reanalysis using a new gridcell dataset for the Northwestern United States.”Journal of Climate,13, 1936–1950.
Wilby RL, Conway D, and Jones PD. (2002a). “Prospects for downscaling seasonal precipitation variability using conditioned weather generator parameters.”Hydrological Processes,16, 1215–1234.
Wilby RL, Dawson CW, and Barrow EM. (2002b). “SDSM – a decision support tool for the assessment of regional climate change impacts.”Environmental Modelling & Software, 17(2), 145–157.
Wilby RL, Hay LE, and Leavesley GH. (1999).”A comparison of downscaled and raw GCM output: implications for climate change scenarios in the San Juan River basin, Colorado.”Journal of Hydrology,225, 67–91.
Wilby RL, and Wigley TML. (1997).”Downscaling general circulation model output: a review of methods and limitations.”Progress in Physical Geography,21, 530–548.
Wilby RL.,(1998).”Statistical downscaling of daily precipitation using daily airflow and seasonal teleconnection indices.”Climate Research,10, 163–178.
Yonas B. Dibike, and Paulin Coulibaly (2005).”Hydrologic impact of climate change in the Saguenay watershed: comparison of downscaling methods and hydrologic models.”,Journal of Hydrology,307, 145–163.
Yonas B. Dibike, Slavco Velickov, Dimitri Solomatine, and Michael B. Abbott (2001).”Model induction with support vector machines:Introduction and applications.”Journal of Computing in Civil Engineering, 208–216.
Yu, X., and S. Y. Liong (2004).”Forecasting of Chaotic Hydrological Time Series with Ridge Linear Regression in Feature Space.”In: Proceedings of the 6th International Conference on Hydroinformatics,S. Y. Liong, K. K. Phoon, and V. Babovic, eds., (Singapore, 21–24 June 2004), World Scientific Publishing Co., Singapore.
Yu, X., S.Y. Liong, and V. Babovic (2002).”Hydrologic Forecasting with Support Vector Machine Combined with Chaos-Inspired Approach. “ Neurocomputing In: Hydroinformatics 2002: Proceedings of the 5th International Conference on Hydroinformatics,(Cardiff, U.K., 1–5 July 2002), IWA Publishing, London, U.K., 764–769.