China's Coastal Typhoon Disaster Assessment Based On Google Earth Engine

China is one of the countries that most severely affected by typhoons.Along with the rapid socio-economic development of China and the high concentration of social wealth and population in coastal areas in recent years,the economic losses caused by typhoons to China have been increasing.On the other hand,global climate change has led to changes in the characteristics of typhoon activity,which indirectly affects the changing risk pattern of typhoon disasters and also poses new challenges for China to cope with typhoon disasters.Therefore,achieving objective and quantitative disaster assessment of typhoon disasters will help to formulate disaster prevention and mitigation policies and delineate loss areas.It is important to reduce the loss of people's lives and properties and to determine the constraining effect of disasters on national economic development.Currently there is still much room to improve the typhoon disaster assessment in China.The traditional disaster assessment is mainly done by grassroots units,which causes many problems such as irregular statistics,and double counting.This affects the government to access disaster information timely and implement reasonable emergency response when disasters occur.In response to these problems,many studies have used new methods to explore their practical application in the disaster assessment process.However,the scope of these studies is often limited to provincial and municipal administrative regions,and cannot cover the entire typhoon affected areas,which is difficult to obtain more comprehensive disaster information about typhoon disasters.In addition,GIS has been fully developed in recent years as a technical tool for practicing typhoon disaster assessment methods.However,most typhoon disaster assessment systems are limited to the disaster assessment of historical periods and cannot perform real-time disaster assessment,making it difficult to give full play to the value and significance of typhoon disaster assessment.Therefore,the traditional disaster statistical methods and application platforms can hardly meet the real requirements of typhoon disaster response due to the lack of global and time-sensitive nature.There is also a need to explore more reasonable and effective methods and applications for typhoon disaster assessment.In this paper,Google Earth Engine,which is a remote sensing big data platform,is used to combine meteorological satellite data,typhoon data and exposed asset data for typhoon impact characterization,disaster assessment and overall design of the assessment system.The specific research contents and results are as follows:(1)This study analyzed the spatial and temporal distribution patterns and long-term trends of typhoon impact characteristics in China from 1981 to 2019 using information such as satellite meteorological data and typhoon data.The results show that the number of typhoons affecting China is decreasing;The annual area affected by typhoons and annual precipitation of typhoons in China are increasing;The future generation time of typhoons in China is mainly clustered in August;The spatial trends of wind and rain impact factors of typhoons are shifting northward and westward respectively.The clustering of typhoon generation time,the expansion of risk range and the increase of typhoon precipitation will probably aggravate the typhoon disaster losses in the future.(2)In this study,we combined the random forest algorithm and multi-source data to build a typhoon disaster assessment model and assessed the typhoon disaster losses from 2000 to 2019.The results show that random forest can achieve good fitting effect in disaster assessment.In the past 20 years,the affected farmland area,affected population and direct economic losses caused by typhoons showed an increasing trend,with the largest growth trend in Southeast coastal areas.In the future,the government needs to focus on typhoon disaster prevention in the economically developed southeastern coastal provinces,and develop a more comprehensive action plan to cope with typhoon disasters to avoid more property damage due to typhoon disasters.(3)This study builds a typhoon disaster assessment system based on the Google Earth Engine platform and disaster assessment model,and establishes a dynamic spatio-temporal database of typhoon disasters.Through the real-time typhoon disaster assessment,the informatization and visualization of typhoon disasters are realized,and the ability to respond to typhoon disasters is greatly improved.