Affiliation:
1. Ocean University of China
Abstract
In order to make up for the lack of natural grade warning, we sought a new method for judging the losses of storm surges. Firstly apply entropy method etc to grade storm surges into 4 levels (mild, moderate, heavy and extra heavy) according to economic loss indices in Zhejiang Province. Then develop BP neural network to forecast losses with the selected indicators of natural, social and economic conditions. Comparing forecast grades with the actual value, we found the accuracy of grade prediction is 80%. It shown the grading results and predicting method are reliable and could be used for the grades of economic losses forecast of storm surges in future.
Publisher
Trans Tech Publications, Ltd.
Reference2 articles.
1. Paolo Antonio Pirazzoli, Flood threat anomaly for the low coastal areas of the English Channel based on analysis of recent characteristic flood occurrences, Ocean Dynamics, No. 57, 2007, pp.501-510.
2. Wang Shengan; You Xiaomin and Li Manqiu, A field experiment of real-time remote monitoring system for nearshore sea wave, storm surge and tsunami and its application, Journal of Tropical Oceanography, No. 28, 2009, pp.29-33.