摘要
本文應用具AR(1)誤差之長時間追蹤資料迴歸模型分析中央氣象局收集到在台北和台南測站自1961年至1994年所觀測到之145個颱風，及台北測站自1995年至2000年所觀測到之31個颱風降雨量的相關資料。與過去的方法相較，本文考慮的長時期追蹤資料迴歸模型對颱風進入北緯22度至26度和東經120度至125度間某時段後之降雨量預測與實際降雨量間之均方誤差較小，相關係數較高。另外本文並嘗試考慮降雨量之區間預測，根據上述資料所得預測區間亦有令人滿意的準確度。

Abstract
We employ the regression models for longitudinal data with AR(1) error vectors to analyze the typhoon rainfall data at Taipei. The data were collected by the Central Weather Bureau that contain the 145 typhoons data from 1961 to 1994 measured at Taipei and Tainan, respectively. Compared with the existing methods, the regression models for longitudinal data yielded smaller mean squared errors and longer correlation coefficients between predicted rainfalls and the real observations. In addition, we also try to make predictive intervals for the rainfalls and the results are quite accurate.

參考文獻
[1] Casella, G., and Berger, R.L. (2001). Statistical Inference (2nd Ed.). Duxbury.
[2] Collins, L.M., and Horn, J.L. (eds.) (1991). Best Methods for the Analysis of Change. Washington, DC: America Psychological Association.
[3] Cox, D.R., and Oakes, D. (1984). Analysis of Survival Data. Chapman and Hall, London.
[4] Dale, A., and Davies, R.B. (eds.) (1994). Analyzing Social and Political Change: A Casebook of Methods. London: Sage.
[5] Daskalakis, C., Laird N.M., and Murphy, J.M. (2002). “Regression Analysis of Multiple-Source Longitudinal Outcomes: A ‘stirling country’ Depression Study.” American Journal of Epidemiology, 155(1), 88-94.
[6] Dempster, A.P., Laird, N.M., and Rubin, D.B. (1977). “Maximum Likelihood from Incomplete Data via the E-M Algorithm.” Journal of the Royal Statistical Society (Ser. B), 39, 1-38.
[7] Diggle, P.J. (1990). Time Series: a Biostatistical Introduction. Oxford University Press, Oxford.
[8] Dutilleul, P., and Pinel-Alloul, B. (1996). “A Doubly Multivariate Model for Statistical Analysis of Spatio-Temporal Environmental Data.” Environmetrics, 7, 551-565.
[9] Geisser, S. (1981). “Sample Reuse Procedures for Prediction of the Unobserved Portion of a Partially Observed Vector.” Biometrika , 68, 243-250.
[10] Gottman, J.M. (ed.) (1995). The Analysis of Change. Mahwah, NJ: Lawrence Erlbaum.
[11] Heitjan, D.F., and Sharma D. (1997). “Modelling Repeated-Series Longitudinal Data.” Statistics in Medicine, 16, 347-355.
[12] Jennrich, R.I., and Schluchter, M.D. (1986). “Unbalanced Repeated-
Measures Models with Structured Covariance Matrices.” Biometrics,42, 805-820.
[13] Laird, N.M., and Ware, J.H. (1982). “Random Effects Models for Longitudinal Data.” Biometrics, 38, 963-974.
[14] Lee, J.C. (1988). “Prediction and Estimation of Growth Curve with Special Covariance Structures.” Journal of the American Statistical Association, 83, 432-440.
[15] Lee, J.C., and Geisser, S. (1975). “Applications of Growth Curve Prediction.” (Ser. A), 37, 239-256.
[16] Louis, T.A., and Spiro, A. Ⅲ (1984). “Fitting First-Order Autoregressive Models with Covariates.” manuscript submitted for publication.
[17] Reinsel, G.C. (1984). “Estimation and Prediction in a Multivariate Random Effects Generalized Linear Model.” Journal of the American Statistical Association, 79, 406-414.
[18] Ware, J.H. (1985). “Linear Models for the Analysis of Longitudinal Studies.” The American Statistician, 39,95-101.
[19] 王時鼎、顏清連、陳泰然與謝信良 (1986). 台灣颱風降雨特性及其預報研究(三)。國家科學委員會防災科技研究報告 74-51，152頁。
[20] 吳宗堯與謝信良 (1990). 現有颱風預報成果作業化之研究(三)。國家科學委員會防災科技研究報告 79-37，137頁。
[21] 李昀寰 (2001). 颱風降雨量與風速之統計預測。國立中央大學統計研究所碩士論文。
[22] 林建發 (2000). 颱風降雨量之統計迴歸預測。國立中央大學統計研究所碩士論文。
[23] 許佳容 (2002). 台灣地區颱風降雨量預測之統計方法探討及校驗。國立中央大學統計研究所碩士論文。
[24] 葉天降、吳石吉與謝信良 (1999). 簡單統計方法於台灣地區颱風降水預測之研究(一)預測方法與台北颱風降水之預測校驗。大氣科學，27，4，395-412。
[25] 葉天降、樊采虹與李昀寰 (2001). 台灣地區颱風降水之迴歸預測(一) 台北颱風降水線性迴歸預測模式。大氣科學，29，1，77-96。
[26] 魏元恒、謝信良與林民生 (1971). 颱風特性與台灣與量之研究。氣象學報，17，第三期，1-17。