Hazard Assessment Of Moraine-dammed Lake Outbursts In The Himalayas, Tibet And The Propagating Numerical Simulation

dam height, the back water slope of the dam and average particle size of moraine deposit embodies moraine dam, four characteristics of heat-and-water combination (hot and humid, dry, wet and warm) (dry) concrete embodiment of climate characteristic. According to the above five characteristics of the concrete embodiment of induction summary form moraine dam lake outburst risk assessment system, namely, the glacier area, glacier affect zone longitudinal slope, distance from glacier tongue to glacier lake, glacier crack development level, lake area, crestwidth, dam height, the back water slope of the dam, average particle size of moraine deposit and heat and water combination,10 indicators. Using rough sets attribute reduction analysis found that selected 10 indicators can be good for the ice Qi lake outburst risk assessment.5. This chapter through the application of artificial neural network method based on the NeuroSolution software emulation multilayer neural modules 5, selecting 30 moraine-dammed lake samples in research area as the training sample of model is established, According to certain rules, Evaluation index dimensionless processing, and then, input neural network software in the form of TXT format, it is concluded that moraine-dammed lake outburst risk assessment neural network model suitable for this study area. Using the neural network model with trained fixed weights and the gradient to study area to assess the hazard of Zuolongma lake, located in NieLaMuXian county, it is concluded that predictive value.for risk prediction is 0.5250 According to risk hierarchies, we can determine the hazard level of Zhuolongma to high risk. Prediction result is consistent with on-the-spot investigation result, It is well proved that neural network model is applicable for hazard assessment of moraine-dammed lake outburst, also further proof that the rationality of the evaluation index system setted up.The new ideas is proposed that the sensitivity analysis of various evaluation indexes should be targeted to specific hazard assessment in a concrete moraine-dammed lake, and to analyse sensitivity of evaluation indexes in outburst hazard assessment of Zhuo longma lake for example, we found the indexes of height and glacier area is very sensitive, and glacier affect zone longitudinal slope and lake area less sensitive. So it is necessary to focus on monitoring glacier area changing of Zhuo longma, also need to monitor lake area changing.6. Useing FLDWAV model to simulate the SangWangCuo lake outburst shows:lake outburst floods will continue to 35 hours,1.0 hours later will reach the maximum flow 1.0×104m3/s, eventually burst depth, up-width and down-width is respectively for 40m, 185m and 58m. From breach location to jingzi town.the distance is 125km, the peak flow will last for 5.9 hour. In the 54km place of downstream, the NeiRuDui county, velocity of flow wave diminishes quickly; flood level drops from 40m above of upstream to 5m below near the village of Baidun 67km place.