由於台灣區域環境的特性，地震、颱風及豪雨為經常侵襲台灣的天然災害，近年來侵襲台灣地區最嚴重的兩個天然災害分別為1999年921地震與2009年莫拉克颱風，在兩次事件中除建物損毀、崩塌與土石流災害類型外，還發生許多的堰塞湖災害；而在莫拉克颱風期間崩塌所引發的堰塞湖災害更造成民眾嚴重傷亡。堰塞湖災害受限於發生地點、時間、規模與潰決的不確定性，相關的預警、分析與應變對策的擬定也較其他災害來的複雜，成為災害防救領域中所面臨新的災害類型。
過去受限於堰塞湖災害觀測資料的欠缺，而無法針對堰塞湖災害進行完整而詳盡的分析，因而在此領域的研究較少，但近年來隨著觀測技術的精進與災害案例的增多，相關基礎資料的已逐步趨於完整；本研究透過台灣地區近年來的堰塞湖災害案例的資料蒐集與分析，利用堰塞湖形成後對下游水位的影響特性，透過資料分析與統計方法建立堰塞湖發生水位異常值檢測方法；同時利用類神經網路中的倒傳遞網路 (BPNN)配合雨量與水位實測資料，透過訓練、驗證建置旗山溪集水區降雨逕流模式，針對莫拉克颱風事件堰塞湖災害實測水位資料進行比對分析，完成旗山溪上游及小林堰塞湖蓄水體積推估，配合堰塞湖所在位置的地形資料，進行堰塞湖形成後之壩體高度推估與潰壩歷程之重現。
本研究透過堰塞湖下游水位歷線的特性分析，利用水位異常檢測進行堰塞湖形成之判釋及蓄水體積推估，如有堰塞湖位置資訊則可透過數值地形進行堰塞湖壩體高度推估，可於災中即時提供防災資訊或於災後重現堰塞湖災害形成與潰決歷程，有助相關災害防救對策的擬定。
關鍵字：堰塞湖、倒傳遞網路、異常值檢測、降雨逕流

Due to regional environmental characteristics of Taiwan, earthquakes, typhoons and torrential rains frequently result in disasters. In recent years, two of the most serious natural disaster events are the Chi-Chi earthquake in 1999 and typhoon Morakot in 2009. These two events caused building damage, landslides, debris flow, and a lots of landslide dams. During the event of typhoon Morakot, a landslide dam caused serious injuries and deaths. Due to the uncertainty of occurrence location, time, scale and failure type of a landslide dam, preparation and response works, such as early warning, analysis and response plan, related to natural dam disasters are extremely difficult. Therefore, this is another new type of disaster prevention issue we need to face.
Subject to limited observation data, complete and detailed analysis for the landslide dams is impossible. In addition, not many studies in this field are available. In recent years, improved observation techniques and data collection investments, the basic database has gradually become completed. In this study, data analysis approach and statistical method are used to detect the occurrence of a landslide dam using abnormal river water level and a back-propagation neural network (BPNN) model. Rainfall and river water level ddata collected recent years are used to train the BPNN model. This BPNN is considered as the rainfall-runoff model of Chi-Shan river basin. The result of the rainfall-runoff model is compared with the observation water level data for detection of the occurrence of a natural dam and for estimation of the water volume stored behind the dam during a landslide event. This proposed method is applied to a landslide dam event in Chi-Shan river basin during typhoon Morakot. Along with the water storage space information and digital terrain model of the landslide dam, the changes of the dam height during the dam failure process can be evaluated.
This study uses the characteristics analysis of downstream water level hydrograph of a landslide dam to identify the occurrence and the water storage behind the dam. By using the digital terrain model of landslide dam area, the proposed method can estimate the height variation of the dam during the dam failure process. The result of the proposed method can provide critical disaster information during response period and re-build the process of the dam, progress between formation and failure, of the dam after the event. The result of this study provides important information for planning the disaster prevention countermeasure.
Keywords：landslide dam, back-propagation neural network, detection of occurrence of natural dams, rainfall-runoff model

1. 王如意、黃智顯，「坡地集水區類神經網路模式之研究及其應用於水文之預測」，台灣水利，Vol.15，NO.1，pp.7-23，1997。
2. 吳佳竣，「小林村災害歷程重現之探討」，國立成功大學水利及海洋工程學系碩士論文，2010。
3. 李孟武、劉長齡，「類神經網路水庫模糊操作模式之建立應用」，第十屆水利工程研討會，1999。
4. 李德河、謝正倫，「地震發生前後地下水水位異常變化之研究」，台日地震前後地下水異常變化研究國際研討會，2005。
5. 宮崎新二、嶋大尚、菊井稔宏、安藤詳平、鈴木啓介、内田太郎，”河川水位の低下に着目した天然ダム発生検知に関する検討”，日本砂防學會研討會，2013。
6. 張斐章、張麗秋，「類神經網路理論與實務」，臺灣東華書局(初版)，2003。
7. 曾皇銘，「小林村潰壩研究」，國立成功大學水利及海洋工程研究所碩士論文，2010。
8. 游保杉等，「自聯想記憶可適性限性類神經網路在波浪模式之研究」，第八屆水利工程研討會，pp1119-1126，1996。
9. 經濟部水利署，「莫拉克颱風水文分析資料」，2009。
10. 農委會林務局屏東林區管理處，「旗山溪上游堰塞湖緊急處置報告」、2010。
11. 農委會林務局屏東林區管理處，「旗山溪(那瑪夏鄉)、士文溪(春日鄉)二處堰塞湖調查評估及短中長期處理對策研擬計畫」，2011。
12. Abrahart, R. j. and See, L., “Comparing neural network and autoregressive moving average techniques for the provision of continuous river flow forcasts in two contrasting catchments”, Hydrological Processes, 14, 2157-2172, 2000.
13. Abrahart, R.J. See, L. and Kneale, P.E., “New tools for neuro hydrologists: using network pruning and model breeding algorithms to discover optimum inputs and architectures”, In Proceedings of the 3rd International Conference on Geo-computation, University of Bristol, 17-19 September, 1998.
14. Atiya, A. F., El-Shoura, S. M., Shaheen, S. I., and El-Sherif, M. S., “A comparison between neural-network for casting techniques-case study: river flow forecasting”, IEEE Transactions on Neural Networks, 10(2), 402-409, 1999.
15. Baba, K. et al., “Explicit representation of knowledge acquired form plant historical data using neural network”, IJCNN-91, III, 155-160, 1991.
16. Berod, D. D., Singh, V. P., and Musy, A., “A geomorphologic kinematic-wave (GKW) model for estimation of floods from small alpine watersheds”, Hydrological Process., 13, 1391-1416, 1999.
17. Betson, R. P., “What is watershed runoff”, J. Geophys. Res., 69(8),1541-1552, 1964.
18. Blikra, L. H and Nemec, W., “Postglacial colluvium in Western Norway: Depositional processes, facies and palaeoclimatic record”, Sedimentology, 45(5), 909-959, 1998.
19. Bodri L, Cermak V, “Prediction of Extreme Precipitation Using a Neural Network: Application to Summer Flood Occurrence in Moravia”. Advances Eng. Software, 31(2000): 311–321, 2000.
20. Brooks, S. M. and McDonnell, R. A., “Research advances in geocomputation for hydrological and geomorphological modeling towards the twenty-first century”, Hydrol. Process., 14, 1899-1907, 2000.
21. Cameron, D., Kneale, P., and See, L., “An evaluation of a traditional and neural net modeling approach to flood forecasting for an upland catchment”, Hydrological Processes, 16, 1033-1046, 2002.
22. Campolo, M., Andreussi, P., and Soldati, A., “River flood forcasting with a neural network model, water Resources Research”, 35(4), 1191-1197, 1999.
23. Carrivick, J. L., “Dam break - Outburst flood propagation and transient hydraulics: A gosciences perspective”, Journal of Hydrology, 380, 338-355, 2010.
24. Carrivick, J. L., Jones, R., and Keevil, G., “Experimental insights on geomorphological processes within dam break outburst floods”, Journal of Hydrology, 408, 153-163, 2011.
25. Carrivick, J.L., and Rushmer, E.L., “Understanding high-magnitude outburst floods”, Geology Today, 22, 60-65, 2006.
26. Casagli, N., Ermini, L. and Rosati, G., “Determining grain size distribution of material composing landslide dams in the Northern Apennine: sampling and processing methods”, Engineering Geology, 69, 83-97, 2003.
27. Chang, F. J. and Chen Y. C., “Estuary water-stage forecasting by using radial neural network”, Journal of Hydrology, 270, 158-166, 2003.
28. Chang, F. J. and Chen, Y. C., “A counterpropagation fuzzy-neural network modeling approach to real time streamflow predictions”, Journal of Hydrology, 245, 153-146, 2001.
29. Chang, F. J. and Hwang, Y. Y., “A-self-organization algorithm for real-time flood forecast”, Hydrological Processes, 13, 123-138, 1999.
30. Chang, F. J., Liang J. M., and Chen Y. C., “Flood forcasting using radial basis function neural networks”, IEEE Transactions on System, Man, and Cybernetics-Part C: Applications and Reviews, 31(4), 530-535, 2001.
31. Chang, T.-J., Kao, H.-M., Chang, K.-H. and Hsu, M.-H., “Numerical simulation of shallow-water dam break flows in open channels using smothed particle hydrodynamics”, Journal of Hydrology, 408, 78-90, 2011.
32. Chen, C. S., Chou F. N. F. and Chen, B. P. T., “Spatial information-based back-propagation neural network modeling for outflow estimation of ungauged catchment”, Water Resources Management, 24 (14), 4175–4197, 2010a.
33. Chen, C.-S., Chen, B. P.-T., Chou, F. N.-F. and Yang, C.-C., “Development and application of a decision group back-propagation neural network for flood forecasting”, Journal of Hydrology, 385(1-4), 173-182, 2010b.
34. Cheng, S. P., “The oral history of a disaster at the Hsiaolin Village”. National Taiwan Museum, Taipei, 246, 2010.
35. Chiang, Y. M., L. C. Chang and F. J. Chang, “Comparison of static-feedforward and dynamic-feedback neural networks for rainfall-runoff modeling”. Journal of Hydrology 290: 297-311, 2004.
36. Clark, C. O., Storage and the unit hydrograph, Tran., ASCE, 110,No. 2261, 1419-1446, 1945.
37. Costa, J. E. and Schuster, R. L., “The formation and failure of natural dams, Geological Society of America Bulletin”, 100, 1054-1068, 1988.
38. Coulibaly, P., Anctil, F., and Bobee, B., “Daily reservoir inflow forecasting using artificial neural networks with stopped training approach”, Journal of Hydrology, 230, 244-257, 2000a.
39. Coulibaly, P., Anctil, F., Aravena, R. and Bobee, B., “Artificial neural network modeling of water table depth fluctuations”, Water Resources Research. 37, 885–896, 2001.
40. Coulibaly, P., Anctil, F., Rasmussen, P., and Bobee, B., “A recurrent neural networks approach using indices of low-frequency climatic variability to forecast regional annual runoff”, hydrological Processes, 14, 2755-2777, 2000b.
41. Davis, W. M., “On the classification of lake basins: Proceedings”, Boston Society of Natural History, 21, 315-381, 1882.
42. Dong, J. J., Li, Y. S., Kuo, C. Y., Sung, R. T., Li, M. H., Lee, C. T., Chen, C. C., and Lee, W. R., “The formation and breach of a short‐lived landslide dam at Hsiaolin village”, Taiwan - Part I: Post-event reconstruction of dam geometry, Engineering Geology, 123(1-2), 40-59, 2011.
43. Evans, S. G. and Hungr, O., “The assessment of rockfall hazard at the base of talus slopes”, Canadian Geotechnical Journal, 30, 620–636, 1993.
44. Feng, Z. Y., “The seismic signatures of the 2009 Shiaolin landslide in Taiwan”, Natural Hazards and Earth System Science, 11 (5), 1559-1569, 2011.
45. Frankel, D. et al. “Use of neural networks to predict lighting at kennedy space center,” IJCNN-91, I, 319-324, 1991.
46. Fread DL,“BREACH: an erosion model for earthen dam failures”, National Weather Service, Office of Hydrology, Silver Spring, MD,1988.
47. French, M. N., Krajewski, W. F., and Cuykendall, R. R., “Rainfall forecasting in space and time using a neural network”, Journal of Hydrology, 137, 1-31, 1992.
48. Garcia-Navarro P. and Villanueva, A. F. I., “Dam-break flow simulation: some results for one-dimensional models of real cases”, Journal of Hydrology, 216, 227-247, 1999.
49. Golob, R, Stokelj, T., and Grgic, D., “Neural-network-based water inflow forecasting, Control Engineering Practice”, 6, 593-600, 1998.
50. Gupta, V. K., Waymire, E., and Wang, C. T., “A representation of an instantaneous unit hydrograph from geomorphology”, Water Resources Research, 16(5), 855-862, 1980.
51. Ham, F. M. and I. Kostanic., “Principles of neurocomputing for science & engineering”, McGraw-Hill, New York, 2001.
52. Hornik, K., Stinchcombe, M. and White, H., “Multilayer feedforward networks are universal approximators”, Neural Networks, 2, 359-366, 1989.
53. Horton, R. E., “The role of infiltration in the hydrologic cycle”, Am. Geophys. Union, 14, 446-460, 1933.
54. Hsu, K-L., Gupta, V.K., Sorooshian, S., “Artificial neural network modelling of the rainfall–runoff process”, Water Resources. Res. 31 (10), 2517–2530, 1995.
55. Hsu, Kuo-Lin, “Application of a recurrent neural network to rainfall-runoff modeling,” Annual Water Resource Planning and Management Conference, 1997.
56. Hutchinson, G. E., “A treatise on limnology”, Vol.1, New York, John Wiley & Sons, 1015, 1957.
57. Kagle, B. J. et al. “Multi-fault diagnosus of electronic circuit boards using neural networks”, IJCNN-90, II, 197-202, 1990.
58. Khalil, M., Panu, U. S., and Lennox, W. C., “Groups and neural networks based streamflow data infilling procedures”, Journal of Hydrology, 241, 153-176, 2001.
59. Kim, G. and Barros, A. P., “Quantitative flood forecasting using multisensor data and neural networks”, Journal of Hydrology, 246, 45-62, 2001.
60. Kimoto, T., and Asakawa, K. “Stock market predication system with modular networks”, IJCNN-90-Wash., I, 1-6, 1990.
61. Kuo, Y.-S., Tsai, Y-J, Chen, Y.-S.m, Shieh, C.-L., Miyamoto, K. and Itoh, T., “Movement of deep-seated rainfall-induced landslide at Shiao Lin Village during Typhoon Morakot”, Landslides. (March, 2012 published online, DOI 10.1007/s10346-012-0315-y), 2012.
62. Kwok, T. Y. and Yeung, D. Y., “Constructive algorithms for structure learning in feedforward neural network for regression problem”, IEEE Transactions on Neural Networks, 3, 630-645, 1997.
63. Li, M. H., Sung, R. T., Dong, J. J., Lee, C. T and Chen, C. C., “The formation and breaching of a short lived landslide Dam at Hsiaolin Village, Taiwan-part II: simulation of debris flow with landslide dam breach”, Engineering Geology, 123(1-2), 60-71, 2011.
64. Liong, S. Y., Lim, W. H., Kojiri, T., and Hori, T., “Advance flood forecasting for flood stricken Bangladesh with a fuzzy reasoning method”, Hydrological Processes, 14, 431-448, 2000.
65. Lo, C. M., Lin, M. L., Tang, C. L. and Hu, J. C., “A kinematic model of the Hsiaolin landslide calibrated to the morphology of the landslide deposit, Engineering Geology”, 123, 22-39, 2011.
66. Luk, K. C., Ball, J. E., and Sharma, A., “An application of artificial neural networks for rainfall forecasting”, Mathematical and Computer Modelling, 33, 683-693, 2001.
67. WS McCulloch, W Pitts, “A logical calculus of the ideas immanent in nervous activity, The Bulletin of Mathematical” , Volume 5 Issue 4, pp 115-133, December 1943.
68. Meyer, W., Schuster, R. L. and Sabol, M. A., “Potential for seepage erosion of landslide dam”, Journal of Geotechnical Engineering (ASCE), 120(7), 1211-1229, 1994.
69. Minns, A.W., Hall, M.J., “Artificial neural network as rainfall–runoff models”, Hydrological Science-Journal 41 (3), 399-419, 1996.
70. Mizuyama, T. Satohuka, Y., Ogawa, K. and Mori, T., “Estimating the Outflow Discharge Rate from Landslide Dam Outbursts” , Disaster Mitigation of Debris Flows, Slope Failures and Landslides, 1, 365-377, 2006.
71. Muftuoglu, R.F., “Monthly runoff generation by non-linearmodels”, Journal of Hydrology, 125, 277-291, 1991.
72. Nash, J. E. (1957). “The form of instantaneous unit hydrograph,” Intl. Assoc. Sci. Hydrology, Pub.45, Vol.3, 114-121.
73. Nemec, W., and Kazanci, N., “Quaternary colluvium in west-central Anatolia: sedimentary facies and palaeoclimatic significance”, Sedimentology, 46(1), 139-170, 1999.
74. Prestininzi, P., “Suitability of the diffusive model for dam break simulation: Application to a CADAM experiment”, Journal of Hydrology, 361, 172-185, 2008.
75. Rodriguez-Iturbe, I., and Valdes, J. B., “The geomorphologic structure of hydrologic response”, Water Resources Research, 15(6), 1409- 1420, 1979.
76. Sajikumar, N., and Thandaveswara, B.S., “A non-linear rainfall–runoff model using an artificial neural network”, Journal of Hydrology, 216(1-2), 32-55, 1999.
77. Shamseldin, A. Y., “Application of neural network technique to rainfall– runoff modelling”, Journal of Hydrology, 199, 272-294, 1997.
78. Sherman, L. K., “Streamflow from rainfall by the unit-graph method”, Eng. New-Rec., 108, 501-505, 1932.
79. Shieh, C. L., Wang, C. M., Lai, W.C., Tsang, Y. C. and Lee, S. P., “The composite hazard resulted from Typhoon Morakot in Taiwan”, Journal of the Japan Society of Erosion Control Engineering, 62(4), 61-65, 2009.
80. Singh, V. P. and Snorrason, A., “Sensitivity of outflow peak and flood stage to the selection of dam breach parameters and simulation models”, Journal of Hydrology, 68(1-4), 295-310, 1984.
81. Swanson, F. J., Oyagi, N. and Tominaga, M., “Landslide dam in Japan, Landslide Dam: Processes Risk and Mitigation”, Geotechnical Special Publication (ASCE), 3, 131-145, 1986.
82. Takahashi, T. and Kuang, S.F., “Hydrograph prediction of debris flow due to failure of landslide dam”, Annuals of Disaster Prevention Research Institute, Kyoto University, 31(2), 601–615, 1988.
83. Tingsanchali, T. and Gautam, M. R., “Application of tank, NAM, ARMA and neural network models to flood forecasting”, Hydrological Processes, 14, 2473-2487, 2000.
84. Tsai, C. P. and Lee, T. L., “Back-propagation neural network in tidal-level forecasting”, Journal of Waterway Port Coastal and Ocean Enginggeing, July/August, 195-202, 1999.
85. Tsou, C. Y., Feng, Z. Y. and Chigira, M., “Catastrophic landslide induced by Typhoon Morakot, Shiaolin, Taiwan”, Geomorphology, 127(3-4), 166-178, 2011.
86. Waythomas, C.F.,” Formation and failure of volcanic debris dams in the Chakachatna River valley associated with eruptions of the Spur volcanic complex”, Alaska, Geomorphology, 39, 111-129, 2001.
87. Wen, C. G., and Lee, C. S., “A neural network approach to multiobjective optimization for water quality management in a river basin”, Water Resources Research, 34(3), 427-436, 1998.
88. Whitely, R., “Approximate confidence intervals for design floods for a single site using a neural network”, Water Resources Research, 35(1), 203-209, 1999.
89. Wooding, R. A., “A hydraulic model for the catchment-stream problem”, I. Kinematic-wave theory, J. Hydrol., 3(3), 254-267, 1965.
90. Xu, Z. X., and Li, J. Y., “Short-term inflow forecasting using an artificial neural network model”, Hydrological Processes, 16, 2423-2439, 2002.
91. Yoon, Y. et al., “A connectionist expert system for dermatology diagnosis”, Expert Systems, 23-31, 1990.
92. Zealand, C. M., Burn, D.H., and Simonovic, S. P., “Short term streamflow forecasting using artificial neural networks”, Journal of Hydrology, 214, 32-48, 1999.