水庫為台灣重要的水資源調配設施，也是幫助防範水文災害的重要緩衝器，近年來因為經濟蓬勃的發展導致民生與工業用水需求逐年遽增，石門水庫運用相當頻繁，造成相關管理單位之供水壓力尤其是乾旱時期大為增加，倘若能事先掌握未來入流量狀況，便能提供決策者進行供水操作決策與乾旱預警之參考。
目前中央氣象局所提供氣候統計降尺度預報產品，為於每月月底預報未來五個月逐月雨量以及溫度狀態，本研究主要目的即為結合此預報系統，探討其可行之使用方法，並透過氣象資料產生器以繁衍未來五個月可能的日雨量及日溫度資料，投入水文模式進行集水區流量推估，進而提供機率預報和定量預報兩種流量預測資訊，且評估兩預報方法之經濟效益。
本研究利用最大機率法則、機率加權法則、降尺度偏差修正值三種氣候統計降尺度預報產品取樣策略，以氣象資料合成模式(WGEN, Tung and Haith, 1995)繁衍水文模式所需之日溫度及日雨量資料，並以技術得分(RPSS、LEPS、MSE)評估流量機率預報能力，結果顯示三種取樣策略皆有大於氣候預報之能力，其中以最大機率法則下有最佳的流量預報能力。而機率和定量兩種流量預報方法亦皆有高於氣候預報的能力，其兩者於六月至十月評估結果中有相似的經濟效益，但在一月至五月評估結果則是以定量預報於實際入流量發生偏低區間時有較廣的經濟效益範圍；以機率預報於實際入流量發生偏高區間時中有較廣的經濟效益範圍。

Resources in Taiwan not only are impotant for water resources management, but also paly as retention measures against flooding. In recent years, the need for domestic and industrial water have increased rapidly because of economic vigorous development, which result in rising stress of water supply especially in drought periods. Therefore, if reservoir inflows can be quantitatively forecasted beforehand, it will be helpful for issuing drought wqrning and making properly decision for water allocations.
The Central Weather Bureau (CWB) issued short-term climate forecasts by statistical downscaling for precipitation and temperature with lead time of 5 months in a 1-month moving window. The objective of this study is to apply the short-term climate forecasts by integrating with a weather generator and a watershed hydrological model to predict inflows of the Shihmen Reservoir with the maximum lead time of 5 months. Both probabilistic flow forecasts and deterministic flow forecasts were produced in this approach, as well as the associated potential economic values of two flow forecasts.
The sampling techniques, including maximum probability, weighted probability, and bias correction probability, were applied to retrieve monthoy mean values of precipitation and temperature from the climate forecast. Then a weather generator was applied to generate daily temperature and precipitation to drive a hydrological model for inflow predictions of the Shimen Reservoir. The skill scores (RPSS, LEPS and MSE) of three sampling results were all greater than climatology skill. The maximum probability approach has the highest predictive ability. Results of both probabilistic flow forecasts and deterministic flow forecasts are also greater than climatology skill, and show certain economic values from June to October. From January to May, the deterministic flow forecasts possess greater economic benefits than that of the probabilistic flow forecasts for cases of observed inflows at blow normal outlooks; while, the probabilistic flow forecasts possess greater economic benefits that than of the deterministic flow forecasts for cases of observed inflows at above normal outlooks.

1.Clark, M.P. and Hay, L.E., 2004. Use of medium-range
numerical weather prediction model output to produce
forecasts of streamflow. J.Hydrometeorol. 5, 15–32.
2.Chu, .J.L., Kang, .H.G., Tam, C.Y., Park, C.K., Chen,
C.T., 2008. Seaonal forecast for local precipitation
over northern Taiwan usingstatistical downscaling. J.
Geophys. Res. VOL.113, D12118,doi: 10.1029/2007JD009424.
3.Ghile, Y.B. and Schulze, R.E., 2009. Use of an Ensemble
Re-ordering Method for disaggregation of seasonal
categorical rainfall forecastsinto conditioned ensembles
of daily rainfall for hydrological forecasting. J.
Hydrology. 371, 85-97.
4.Gobena, A.K. and Gan, T.Y., 2010. Incorporation of
seasonal climate forecasts in the ensemble streamflow
prediction system. J. Hydrology. 385, 336-352.
5.Huang, W.C. and Chou, C.C., 2005. Drought early warning
system in reservoir operation: Theory and practice.
Water Resour. Res., Vol. 41, W11406,
doi:10.1029/2004WR003830.
6.Kim, J.W., Chang, J.T., Baker, N.L., Wilks, D.S., Gates,
W.L., 1984. The statistical problem of climate
inversion: determination of therelationship between
local and large-scale climate. Monthly Weather Review.
112 (10), 2069-2077.
7.Landman, W. A., S. J. Mason, P. D. Tyson, W. J. Tennant,
2001,Statistical downscaling of GCM simulations to
streamflow, J. Hydrology. Vol.252, pp.221-236.
8.Li, H., Luo, L., Wood, E.F., Schaake, J., 2009. The role
ofinitialconditions and forcing uncertainties in
seasonal hydrologic forecasting. J. Geophys. Res. 114,
D04114, doi:10.1029/2008JD010969.
9.Murphy, J., 1999. An Evaluation of Statistical and
Dynamical Techniques for Downscaling Local Climate. J.
Climate 12, 2256–2284.
10.Mason, S.J. and Graham, N.E., 1999. Conditional
probabilities,relative operating characteristics, and
relative operating levels. Vol. 14, 713-725.
11.Maurer, E.P. and Hidalgo, H.G., 2008. Utility of daily
vs. monthly large-scale climate data: an
intercomparison of two statistical downscaling methods.
Hydrology and Earth System Sciences. 12(2), 551-563.
12.Muluye, G.Y., 2011. Implications of medium-range
numerical weathe model output in hydrologic
applications: Assessment of skill and economic value.
J. Hydrology. 400, 448-464.
13.Potts, J.M., Folland, C.K., Jolliffe, I.T., Sexton, D.,
1996. Recised LEPS scores for assessing climate model
simulations and long-range forecasts. J. Climate. 9, 34-
43.
14.Pan, J., and H. van den Dool, 1998. Extended-range
probability forecasts based on dynamical model output.
Wea. Forecasting, 13, 983–996.
15.Richardson, D.S., 2000. Skill and economic value of the
ECMWF ensemble prediction system. Quart. J. Roy.
Meteorol. Soc. 126,649-668.
16.Roulin, E., 2007. Skill and relative economic value of
medium-range hydrological ensemble predictions. Hydrol.
Earth Syst. Sci. 11, 725–737.
17.Storch, H.V., Zorita, E. and Cubasch U., 1993.
Downscaling of global climate change estimates to
regional scales an application to Iberian rainfall in
wintertime. J. Climate. 6, 1161-1171.
18.Shi, X., Wood, A.W. and Lettenmaier, D.P., 2008. How
essentialis hydrologic model calibration to seasonal
streamflow forecasting. J.Hydrometeorol. 9,1350–1363.
19.Tocci, C.E.M., Collischonn, W., Clarke, R.T., Paz,
A.R., Allasia, D., 2008. Short- and long-term flow
forecasting in the Rio Grande watershed (Brazil).
Atmos. Sci. Lett. 9, 53–56.
20.Ward, M.N. and Folland, C.K., 1991. Prediction of
seasonal rainfallin the Nordeste of Brazil using
eigenvectors of sea-surface temperature. Int. J.
Climatol. 11, 711-743.
21.Wilks, D.S., 2006. Statistical Methods in the
Atmospheric Sciences. Academic Press. Elsevier.
Amsterdam. 627 pp.
22.WMO, 2002: Standardized verification system (SVS) for
long-range forecasts (LRF). New attachment II-8 to the
Manual on the GPDS (WMO-No. 485). Available online at
http://www.bom.gov.au/wmo/lrfvs/Attachment_II-8.doc。
23.Werner, K., Brandon, D., Clark, M., Gangopadhyay, S.,
2005.. Incorporating mediumrangenumerical weather model
output into the ensemble streamfolw prediction system
of the National Weather Service. J. Hydrometeorol. 6,,
101-114.
24.楊道昌、陳姜琪、吳雷根、游保杉，2003，結合水文模式與氣象
資料預報長期逕流量之研究，農業工程學報，第49卷，第1期，
第23-33頁。
25.田維婷，2003，氣候變遷對台灣地區地表水文量之影響，中央大
學水文與海洋科學研究所碩士論文。
26.吳雷根，2004，曾文水庫枯水期長期入流量預測之研究，國立成
功大學水利及海洋工程研究所碩士論文。
27.游保杉，謝龍生，連宛渝，2006，台灣南部地區長期雨量變動與
乾旱特性變化趨勢，國家災害防救科技中心，災害防救電子報第
14期。
28.游保杉，2007，應用氣象展望資料於長期流量預報(二)，行政院
國家科學委員會專題研究計畫報告。
29.胡志文，蕭志惠，童雅卿，任俊儒，鄭凱傑，黃文豪，施宇晴，
施景峰，莊穎叡，賈愛玫，2008，中央氣象局動力統計氣候預報
系統簡介，天氣分析與預報研討會。
30.魏綺瑪，2009，利用統計降尺度法推估石門水庫集水區未來情
境降水研究，國立成功大學水利及海洋工程研究所碩士論文。
31.陳正達，2009，台灣與東亞地區之月、季定量降水估計技術研發
與應用，交通部中央氣象局委託研究計畫報告。
32.許勝傑，2009，水庫乾旱風險預警系統之建置-以基隆地區為
例，國立台灣海洋大學河海工程學系碩士論文。
33.林思達，2009，改良GWLF模式應用於翡翠水庫入流量模擬，中央
大學水文與海洋科學研究所碩士論文。
34.蕭震洋，謝寶珊，高丞瑋，林伯勳，鄭錦桐，2009，石門水庫集
水區不同時期崩塌地變遷分析與探討，中國地球物理學會與中華
民國地質學會98年年會暨學術研討會論文摘要集，嘉義，五月，
第HZ04頁。
35.劉正欽，2010，區域氣候場對雨量降尺度之影響，國立成功大學
水利及海洋工程研究所碩士論文。
36.陳孟詩，2010，中央氣象局月季長期天氣展望之預報校驗，天氣
分析與預報研討會。
37.童新茹，2011， 結合季長期天氣預報與水文模式推估石門水庫
入流量，中央大學水文與海洋科學研究所碩士論文。
38.吳阜峻，2011，通用型水資源調配模式之發展與應用-枯水期石
門水庫缺水風險分析，國立交通大學土木工程研究所碩士論文。
39.經濟部水利署水文水資源資料管理供應系統網站
http://gweb.wra.gov.tw/wrweb/。