A Flood Simulation Based On The Types Of Flood Frequencies And The Generalized Flood Risk Analysises With Their Simulations

The flood can give human life and property huge disaster risks, and the root causes of flood risks are the corresponding preventive measures and the scheduling’s uncertainties resulted by the stochastic of the future flood. At present, the main methods to describe the future flood concern the various simulation technologies, which can be utilized to simulate the flood’s main characteristics and correlation of its peak, volume and duration. And then the simulation analyses of the flood risks are launched based on the flood simulation. It is attentioned that scheduling and response measures vary with the flood frequencies, and the subsequent flood risks can be as like as chalk and cheese. Therefore, it is a temporarily neglect but deserving attention problem to maintain the future flood’s frequency while simulating, which can help us to give our subjective initiative a better play, and then improve the effect of our flood risk prevention. However, the current flood risk analyses basically focus on risk probability, which resulting in an embarrassing situation that the floods with a risk rate "once a hundred year" frequently occur in the recent years. It indicates that we should improve narrow-sense flood risk analyses to generalized-sense ones which concern ont only the the flood risk rate but also the depiction of the risk hazard degree and other multiple characteristics of flood risk. The work in this paper can be classified into four respects as the following, which revolve around a more realistic simulation of unknown flood process and more clear analysis of the flood risk.First of all, in order to make the simulated flood used in the flood analysis maintain the future flood’s frequency, an idea of simulating the flood processes according to the flood frequency classification was put forward, and the corresponding simulation framwork was constructed, and then flood process simulation of Three Gorges reservoir was took as an example to test the simulation effect. It means to make the simulated floods maintain the particular distribution of historical flood frequency, and then to avoid the generated floods too average to ensure the pertinence of subsequent risk simulations. The specific implementation method of flood simulation can be summarized into five steps:in the first step, a flood type is simulated according to the historical flood frequency classification and distribution; in the second step, a peak and the related volume are simulated according to the maginal PⅢ distribution and the Coupla function of the simulated type; the third step concerns how to choose a historical flood process which is in the simulated type and whose value of the ratio of flood peak to the volume is most close to the simulated value; and then in the forth step, an amplification method with variable ratios is adopted to simulate the flood process. The simulation effect of flood process in Three Gorges reservoir indicated that, compared with the simulated floods by traditional Coupla technology, the simulated floods in this paper have better retention with respect to the flood frequency. What’s more, because each match in this method does not need to be excuteed in the whole history flood database, but only in a certain kind of flood subdatabase, the calculation of flood process quantity is smaller and the simulation is faster.Secondly, in order to improve the narrow-sense flood risk awkwardness that the floods with a risk rate "once a hundred year" frequently occur in the recent years, a kind of generalized flood risk which can reflect not only the flood risk rate but also the risk severity degree was put forward; in order to apply the adaptive scheduling decision, classifying the possible scheduling decision according to the type of simulated flood frequency was discussed; the simulation framework of the generalized risk analysis by Monte Carlo was researched; and the generalized risks of Three Gorges reservoir in2009were simulated to test the effects of the ideas and simulation of the generalized risk analysis. The specific operation of the process mainly includes four steps. The first step, give the definitions of the flood risk rate and severity degree. The second step, simulate the flood processes by utilizing the method which is put forward in this paper. The third step, classify the possible scheduling decision according to the types of flood frequency and choose the scheduling decision according to the simulated flood frequency type. The forth step, simulate the generalized risk analysis by Monte Carlo. Thirdly, in order to research the risk of reservoir impoundment via the generalized flood risk which can simultaneously depict the flood risk rate and severity degree, a generalized flood regulation risk based on VaR is suggested. The simulation of the generalized flood regulation risk analysis of Three Gorges reservoir by Monte Carlo indicated that:as the impoundment is lower than the expectation, the flood regulation risk can describe the risk rate and severity degree of the failure; as the impoundment is lager than the reservoir capacity, the flood regulation risk can also depict the risk rate and severity degree of the flood. Therefore the flood regulation risk based on VaR is a generalized flood regulation risk. The main implementation process can be classified into four steps. In Step1, the definition of the flood regulation risk based on VaR is given. In Step2, the flood processes is simulated by utilizing the method which is put forward in this paper. In Step3, the possible flood regulations are classified and chosen according to the simulated flood frequency types. In Step4, the generalized regulation risk analysis is simulated by Monte Carlo.The full text was divided into seven chapters, which were relatively independent and mutually relevant yet.In the first chapter, the main research background and significance were introduced, the development history of the flood risk theory and the early classic literature were briefly reviewed, and then the theory development process as well as research achievements of the flood risk were summarized and systematically reviewed.The second chapter focused on two elementary concepts in multivariate stochastic hydrology:P-III distribution and Coupla technology, which provided the theoretical and technical foundation for the follow-up flood process simulation and corresponding risk analysis.In the third chapter, Monte Carlo simulation method and the basic implementation principle by Matlab were mainly studied. It provided a plausible method for the simulation of flood processes and risk analyses.The fourth chapter mainly studied the flood process simulation according to the frequency classification of the historical floods, where the simulation tools consisted of Monte Carlo simulation method and Coupla technology. For the flood process simulation, two important dependent intrinsic variables flood peak and volume were selected. To test the effect of this simulation method, Three Gorges reservoir flood proceses were simulated. The results showed that this method made the flood simulation better in the flood frequency retention and faster in calculation.The fifth chapter studied a generalized flood risk which could depict the flood risk rate and severity degree as well. A well defined generalized flood risk was defined and then the simulation of flood process for possible selection decision were simulated, finally the generalized risk analysis of Three Gorge Reservoir in different flood control scheme and operation rules were simulated. The simulation effect of the generalized flood risk analysis of Three Gorges reservoir indicated that, compared with the narrow-sense flood risk analysis, the generalized flood risk analysis can avoid the embrassing situation that the floods with a risk rate "once a hundred year" frequently occur, and it can help us to understand the scheduling measures more clearly and thus choose the better scheduling mode to minimize the future flood risk degree. As for the Three Gorges reservoir, it can help us understand the flood risk characteristic more fully and clearly, and provide richer arguments to measure or estimate the consequences of elevation or change of flood water level, to give references to take a more appropriate scheduling decision.The sixth chapter used the financial risk analysis tool VaR theory to define a generalized flood regulation risk, which could inegrate the risk of reservoir impoundment and the generalized flood risk. In the generalized flood regulation risk, the risk rate is the probability of the flood, while the risk degree is the corresponding VaR. Taking the flood regulation risk of Three Gorges reservoir as a numerical example, two different type regulations’corresponding risks in2010and Var (0.01) of2011were simulated by Monte Carlo. The results indicated that:the difference between the generalized flood regulation risk based on VaR and the classic flood regulation risk can be comprehended from two aspects. The surface difference lies in the priority order of considering the question:as for the classic flood risk, the flood risk rate is firstly calculated according to the historical flood data, then the scheduling rules are adjusted; as for the generalized flood risk based on VaR, all the generalized flood regulation risks of the combinations of dispatching rules and lood control water levels are firstly appraised according to the simulated floods, then the appropriate combination of regulation and flood control water level is selected. The deeper difference lies in the initiative and passivity of the flood risk control. As for the classic flood regulation risk, the dispatching mode ws passively chosen according to the future or current flood risk rate calculated before. But as for the generalized flood regulation risk based on VaR, the situation is on the opposite:all the dispatching mode were actively appraised by its possible resulting losses in advance, then the strategy was chosen according to the actual ability to bear the risk. This is due to VaR technology which is an important tool in the financial risk analysis to control the risk of a performance. The generalized risk analysis method offered a new perspective for describing flood risks.The seventh chapter made a conclusion and prospect. One, look right in the simulation method for use with Monte Carlo simulation method, with a more realistic simulation of random process of flood, the flood risk analysis of generalized reference value greater; secondly, in flood simulation generation mechanism, look forward to the Coupla technology and seasonal quantile regression combination, make the simulation of flood process the effect is better, thereby improving the generalized risk analysis results.