Assessment of world disaster severity processed by Gaussian blur based on large historical data: casualties as an evaluating indicator

N. Zhang,H. Huang
DOI: https://doi.org/10.1007/s11069-018-3199-x
IF: 3.158
2018-01-01
Natural Hazards
Abstract:Natural and man-made disasters seriously threaten human life. Knowledge about the severity of disasters in general, and the disaster severity of individual countries, is useful in helping to reduce loss of life and economic losses caused by these disasters. In this paper, we provide a Gaussian blur-based method to calculate the average severity of disasters, instead of using the mean or median values as the average severity. This method can partly eliminate the right skewing that is a result of few serious disasters and the left skewing resulting from a great number of small disasters. A new definition of severity based on a natural logarithm is put forward to quantify the severity of all disasters. Droughts, extreme temperatures, and earthquakes are the top three disasters with the highest severity values. Storms have the highest uncertainty, although their severity is low. After analyzing the hazards of countries, China, Indonesia, India, and America were found to be the four highest hazard countries in the world. Finally, we established an annual disaster hazard value per unit area ( H area ) to represent the severity of disasters of countries, taking into account the country’s area. Island countries naturally have high H area , while most of the other high- H area countries lie in Africa.
What problem does this paper attempt to address?