Evaluation of Loss Due to Storm Surge Disasters in China Based on Econometric Model Groups

• Published in: International journal of environmental research and public health 2018-03, Vol.15 (4), p.604
• Main Authors: Jin, Xue, Shi, Xiaoxia, Gao, Jintian, Xu, Tongbin, Yin, Kedong
• Format: Article
• Language: English
• Subjects: Access control; Agricultural economics; Agricultural engineering; China; Coastal zone; Cyclones; Damage assessment; Disasters; Division; Earthquakes; Econometrics; Emergency preparedness; Emergency warning programs; Evaluation; Fisheries; Humans; Hurricanes; Insurance claims; Mathematical models; Model testing; Models, Econometric; Storm surges; Storms; Stormwater; Support Vector Machine; Support vector machines; Water levels; Weather; Wind speed
• ISSN: 1660-4601; 1661-7827; 1660-4601
• DOI: 10.3390/ijerph15040604

Storm surge has become an important factor restricting the economic and social development of China's coastal regions. In order to improve the scientific judgment of future storm surge damage, a method of model groups is proposed to refine the evaluation of the loss due to storm surges. Due to the relative dispersion and poor regularity of the natural property data (login center air pressure, maximum wind speed, maximum storm water, super warning water level, etc.), storm surge disaster is divided based on eight kinds of storm surge disaster grade division methods combined with storm surge water, hypervigilance tide level, and disaster loss. The storm surge disaster loss measurement model groups consist of eight equations, and six major modules are constructed: storm surge disaster in agricultural loss, fishery loss, human resource loss, engineering facility loss, living facility loss, and direct economic loss. Finally, the support vector machine (SVM) model is used to evaluate the loss and the intra-sample prediction. It is indicated that the equations of the model groups can reflect in detail the relationship between the damage of storm surges and other related variables. Based on a comparison of the original value and the predicted value error, the model groups pass the test, providing scientific support and a decision basis for the early layout of disaster prevention and mitigation.