台灣地區的降雨量在時間分配上相當不均，一般5~10月為豐水期，70%~90%的降雨集中在豐水期，而11~4月為枯水期，尤其南部地區超過半年時間幾乎無雨。近年來全球各地的水文狀況變動加劇，多數認為是受到氣候變遷的影響，豐枯水期的變化比起以往更加劇烈，從近年資料可發現臺灣地區每年的降雨天數逐漸減少，但強度及總雨量卻增加，今日從事水資源開發規劃時，必須考慮氣候變遷對水文過程的影響，以掌握未來可能的豐枯變異。
目前對氣候變遷影響的預估分析，一般是利用大氣環流模式求得氣候變遷情況下之平均雨量變化率，再將預估雨量過程輸入降雨－逕流模式，考慮集水區物理特性之地形、地貌、土地利用、河川網路等因素，估算流出水量。但本研究假設地文因素不變，探討使用時間序列的轉換函數法，由蒐集的歷史降雨及水文紀錄，建立雨量及流量過程的線性時間序列模式，據以繁衍氣候變遷情境下可能的集水區流出量過程。
本研究取大甲溪石岡壩水庫集水區為分析案例，將集水區自德基大壩分為上下游兩區，建立雨量和流量相關之多變量時間序列轉換函數(VARMAX)月模式及旬模式。對多年度的水文資料，利用動差法估計模式參數，並比較多變量下之AIC指標，決定各月或旬模式之最適合階次與參數。
選定IPCC第五次評估報告(AR5)之基期時間1986~2005年的歷史資料據以建立模式，再選5種對水資源運用評估較適用的全球環流模式(GCM)推估各月份之雨量改變率，最後比較模式繁衍流量與受氣候變遷影響的2006~2014年歷史流量之間的統計特性，認為發展之模式繁衍成果頗為合理。

In this study, the watershed of the Dajia river was taken as an analysis case.
The watershed was divided into two areas, the upstream and downstream watersheds, and the multivariate time series transfer function (VARMAX) model related to rainfall and flow was established. Because there are many years of hydrological mode data, the moment method is used to estimate the model parameters, and the multi-variable AIC indicators are compared to determine the most suitable order and models for month or ten-day. According to the IPCC Fifth Assessment Report (AR5) was selected for the base period (1986~2005), the five applicable global circulation models (GCM) were used to estimate the monthly rainfall change rate. Finally, comparing the statistical characteristics between historical data and generation data, it is considered that the model generation results are quite reasonable.

1. 涂秀錦、林國峰. (1987)，「混合機率分佈與合成流量之研究」，國立台灣大學土木工程學研究所碩士論文。
2. 孫永信(1988)，「河川旬流量之合成與預測」，國立台灣大學土木工程學研究所碩士論文。
3. 郭振泰、朱文生、徐年盛、陳昶憲、林永禎、孫永信(1991)，「淡水河流域水庫系統即時優選操作模式之發展與應用(三)：小時模式研究」，經濟部水資源局統一規劃委員會，80水利科技八(二)3(10)014第078號。
4.　易任、王如意(1992)，「應用水文學」，國立編譯館。
5.　陳昶憲、李姍燁(1998)，「洪流之多變量時序分析」，中華土木保持學報，29(3)：199-209。
6.　游山峰(1999)，「石門水庫運轉規線之研究」，國立台灣海洋大學河海工程學系碩士論文。
7.　游保衫、王毓麒(2000)，「水庫式電廠可靠發電量不確定之研究」，中興工程科技研究。
8.　溫漢章(2001)，「利用合成流量進行水資源系統分析」，國立台灣大學土木工程學研究所碩士論文。
9.　趙建勛、周乃昉(2015)，「修正分配模式繁衍短延時序率降雨過程之研究」，國立成功大學水利及海洋工程研究所碩士論文。
10.　HARMS, AA; TH Campbell.(1967), An Extension to Thomas-Fiering Model for Sequential Generation of Streamflow, Water Resources Research.
11.　C. K. Young(1968), Discussion of Mathematical Assessment of Synthetic Hydrology’ by N. C. Matalas, Water Resources Research.
12.　T. A. MCMAHON(1971), Application of the Thomas and Fiering Model to Skewed Hydrologic Data, Water Resources Research.
13. Box, G. E. P. and G. M. Jenkins(1976), Times Series Analysis: Forecasting and Control , Holden-Day, Inc., San Francisci.
14. F. Camacho and A. I. McLeod(1987), Multivariate Contemporaneous ARMA model with hydrological applications, Stochastic Hydraulic, vol. 1, pp141-154.
15. H. Raman and N. Sunilkumar(1995), Multivariate Modelling of Water Resources Time Series Using Artificial Neural Networks, Hydrological Sciences Journal, vol. 40.
16. Boughton, W.C.(1999), A daily rainfall generating model for water yield and flood studies. Report 99/9, CRC for Catchment Hydrology, Monash University, Melbourne, 21pp.
17. Chaouche(1999), A. and Parent, E., Bayesian identification and validation of a daily rainfall model under monsoon conditions. Hydrol. Sci. J., 44, 199–220 (in French).
18. R. Srikanthan and T. A. McMahon(2001), Stochastic generation of annual, monthly and daily climate data ：A review, Hydrology and Earth System Sciences, 5(4), 653-670.
19. Deepesh Machiwal and Madank. Jha(2006), Time Series Analysis of Analysis of Hydrologic Data For Water Resources Planning And Management, J.Hydrol. Hydromech., Vol. 54, pp. 237-257.
20. Kenneth Strzepek and Alyssa McCluskey(2007), The Impacts of Climate Change on Regional Water Resources and Agriculture in Africa, Policy Research Working Paper,4290.
21. Julia Piantadois(2009), John Boland, Phil Howlett, Generating Synthetic Rainfall on Various Timescales-Daily, Monthly and Yearly, Vol 14, pp 431-438.
22. Jemma Gornall, Richard Betts(2010), Eleanor Burke, Robin Clark, Joanne Camp,Kate Willett, and Andrew Wiltshire, Implications of climate change for agricultural productivity in the early twenty-first century, Philos Trans R Soc Lond B Bioi Sci, vol. 365.
23. Tze Leung Lai and Ka Wai Tsang(2016), Multivariate Stochastic Regression In Time Series Modeling, Statistica Sinca 26, 1411-1426.
24. J. R. Lamontagne and J. R. Stedinger(2018), Generating Sythetic Streamflow Forecasts with Specified Precision, American Society of Civil Engineers.