Risk Analysis Of Typhoon Rainstorm And Flood Disaster Based On GPM IMERG Data

Typhoons,accompanied by strong winds,storm surges,extreme rainfall and the resultant flash floods and mudslides,are one of the most damaging types of natural disasters around the world.With global warming,the number and proportion of high-intensity typhoons have increased dramatically.China is located in the northwest Pacific region,where typhoons have the highest frequency and intensity.Typhoons frequently cause severe torrential rainfall and flooding in the eastern part of the country,seriously threatening people's property and lives.Current research on typhoons has focused more on the strong winds and their associated hazards,with less attention on the hazards of typhoon flooding.Studies on typhoon precipitation are also mostly limited to individual enents or parts of the country,lacking an overall understanding of the long-term typhoon changes in flooding hazard across China.The successful launch of the new generation of GPM(Global Precipitation Measurement)satellite provide technical support for long-time and large-scale typhoon precipitation studies.The main research contents and conclusions of this paper are as follows:(1)A datdaset of 304 typhoons precipitation events in China from 2000-2019 was constructed based on GPM satellite precipitation products.The overall accuracy of the GPM precipitation products under typhoon condition and their ability to capture the spatial and temporal variability characteristics of typhoon precipitation were evaluated on hourly and daily scales,respectively,in combination rain gague observations,the applicability of mean field bias correction method to GPM precipitation products under typhoon weather was explored.The results show that the GPM can capture the whole typhoon precipitation well at both time scales,and the correlation with the rain gagues is above 0.8.The spatial and temporal variability characteristics of typhoon gagues is above 0.8.The spatial and temporal variability characteristics of typhoon precipitation are better captured at the daily scale(correlation coefficient of 0.72).As rain intensity increases,the GPM's ability to detect heavy rainfall events is reduced by underestimating heavy precipitation.The absolute error in precipitation observations decreases with increasing distance to the typhoon eye;the relative error tends to increase slightly with increasing distance,and is smallest near the eye of typhoon.The segmented mean field bias correction method is better than the overall correction,but the improvement of the overall accuracy is still limited.(2)The spatial and temporal characteristics of typhoon precipitation in China are analyzed from two perspectives:typhoon view and geographical view.The results show that under the typhoon view,the GPM satellite precipitation products can well capture the precipitation characteristics of typhoons at different locations,with high precipitation frequency(0.7-0.9),intensity(100 mm every typhoon)and range(100 km)to the left side of the moving direction or left front of the wind shear direction,and higher risk of heavy rainfall and flooding;compared with the moving direction,the wind shear has more influence on the typhoon precipitation distribution.Under the geographical view,the annual average precipitation of typhoons in China decreases from the southeastern coast to inland areas in the past 20 years,and the highest annual average precipitation is in Hainan,which exceeds 800 mm;other areas with serious hazards are Taiwan Island,Hainan Island,and coastal areas of Guangdong and Guangxi provinces,followed by coastal areas of Fujian and Zhejiang Province;influenced by individual typhoons,some inland areas such as Sichuan,Inner Mongolia,and Heilongjiang have higher precipitation every typhon and hourly average precipitation.The annual number of typhoons in the southeastern coastal areas is much higher than that in the northern coastal areas,with a spatial trend of decreasing from southeast to northeast.The region with the highest number of typhoons per year is Taiwan Island,with an average of more than 10 per year;the northeastern region has the least,with Shandong,Jilin,Liaoning and Heilongjiang having an average of 2 per year.In the past 20 years,typhoons have a tendency to move northward,leading to a decreasing trend in the maximum annual typhoon precipitation in the southeastern coastal region,while there is an increasing trend in the northeastern region,and the risk of typhoon storm flooding in the northeastern region cannot be ignored.(3)The normalized rainfall index ERM of 304 typhoons was calculated to reveal the spatial characteristics of typhoon flood risk in China over 20 years,the typhoon flood risk index was constructed and its practicality was fully verified.The results found that coastal areas are highly susceptible to heavy rainfall and flooding,with Hainan Island having the highest risk of being affected;Guangdong,Guangxi,Hunan,Jiangxi,Zhejiang and Jiangsu have all experienced extreme precipitation exceeding three times the typical heavy precipitation of a two-year local event,bringing serious loss of life and property;inland areas such as Sichuan,Chongqing,Shaanxi and Shanxi generally have a lower risk;the Northeast Plain and North China Plain have a lower risk due to their flat terrain and increased intensity of typhoons occurring in recent years,the risk of flash floods is higher.Several storm flood risk assessment indicators constructed based on ERM have a high correlation with the actual disaster losses of typhoons.Among them,the 90th percentile ERM considering different durations of precipitation in the typhoon life cycle can better reflect the risk of flash flooding and is more consistent with the actual disaster loss.ERM does not depend on ground rainfall stations.It can be obtained conveniently using satellite precipitation,and cover areas with sparse distribution of rain gagues,and is also indicate the relative risk degree of each place in space.