Author
Yang Fei,Meng Jianming
Journal
International Journal of Disaster Risk Science
Paper Keyword
Heat wave,Apparent temperature,Machine learning,Risk assessment,monthly variation
Abstract
Under the background of climate change, the increasingly frequent extreme high temperature events have caused more significant harm to human life security and social and economic development, and become one of the most influential major climate disasters in the world. In particular, the heat wave events in South and Southeast Asia have caused great risks and harm to the development of the region. Based on the real death population in the historical heat wave events, combined with the population data, economic data, infrastructure data related to the event, this study evaluates the heat wave risk in mainland of South Asia and Indochina Peninsula by machine learning model, and makes a quantitative analysis of influencing factors contributions to heat wave risk, and analyzed further the monthly temporal and spatial changes of heat wave indicators from April to September. The studied results show that, the heat wave duration (HWD), heat wave intensity (HWI), heat wave frequency (HWF), and the population of above 65 years old contribute the most to the heat wave risk. The high-risk areas of heat wave disaster in study area have the characteristics of high cohesion and concentration in regional distribution. The high-risk areas are mainly concentrated in the central region of Pakistan and India, northeastern India, the southeast coastal zone of Bangladesh and the southern part of the Mekong River region of Myanmar. The heatwave will become much more in South Asia and Indochina Peninsula, the central and northeastern India rises the fastest. May and July are the months with great heatwave changes. The HWI and HWATmax, show obvious rising changes in July, HWD shows rising changes in May, in the eastern and central plateaus of India. South Asia is suffering from heat wave averaged lasting more than 60 days a year, and the maximum apparent temperature reaching more than 40 ℃. The heat wave risk evaluation with historical events death toll and monthly variations will provide more concrete research references for disaster prevention.
