Flood Risk Analysis Based on Information Diffusion Theory

ABSTRACT

Due to the fact that the flood data series of small drainage basins is relatively short, available data are often not sufficient for flood risk analysis. This presents the problem of risk analysis using very small data samples. One method that can be applied is to regard the available small samples as fuzzy information and optimize them using information diffusion technology to yield analytical results with greater reliability. In this article a risk analysis method based on information diffusion theory is applied to create a new flood risk analysis model. Application of the model is illustrated taking the Jinhuajiang and Qujiang drainage basins as examples. This is a new attempt at applying information diffusion theory in flood risk analysis. Computations based on this analytical flood risk model can yield an estimated flood damage value that is relatively accurate. This study indicates that the aforementioned model exhibits fairly stable analytical results, even when using a small set of sample data. The results also indicate that information diffusion technology is highly capable of extracting useful information and therefore improves system recognition accuracy. This method can be easily applied and the analytical results produced are easy to understand. Results are accurate enough to act as a guide in disaster situations.

Key Words:

• flood
• information diffusion
• risk level
• risk analysis

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (No. 40771044) and the Zhejiang Provincial Science and Technology Foundation of China (No. 2006C23066).