| China is one of the countries most seriously affected by natural disasters in the world because of its vast territory,complex geographical environment,changeable climate and frequent natural disasters.The traditional disaster statistics mainly include the survey and statistics conducted by the investigators,which has a large workload,a long cycle and a high risk coefficient,while the nightlight remote sensing has a strong timeliness,a wide coverage and a short return cycle,can capture the tiny light to reflect the disaster situation dynamic change.Based on the NPP-VIIRS nightlight data,this paper studies the natural disasters of“Typhoon Lekima”and“No.1 flood in Xijiang in 2020”,firstly,the suitability of the model and the scale between the light index and the population and GDP are analyzed.Secondly,the affected area and the affected population are inversed to realize the rapid and accurate assessment of the disaster situation and the judgment of the recovery process,at the same time,a disaster visualization system is constructed based on Arc GIS Online cloud platform.The results show that NPP-VIIRS nightlight remote sensing data can be used for natural disaster assessment.The main research contents are as follows:(1)Model suitability and scale suitability analysis.In this research,the spatial geometric matching of the nightlight remote sensing data is carried out,and the multi-sample region denoising is selected based on the radiation normalization of the statistics method,the average intensity of light,total intensity of light and area of light are analyzed,and the suitability of population and GDP model is evaluated by linear and polynomial models.The results show that the polynomial model is better than the linear model,and Zhejiang chooses the polynomial model of urban population density and light density(R~2>0.89),GDP and total light number(R~2>0.84)to do the inversion,urban population density and light area(R~2>0.75),GDP and light area(R~2>0.94)were used to invert the model in Guangxi.(2)Research on rapid disaster assessment method.Based on the NPP-VIIRS DNB data,the light changes before and after the disaster were identified,the affected area was obtained by the difference,and the affected population was quantitatively assessed,and the severity of the disaster was determined according to the affected area and the number of affected population.The results show that the relative error between the assessment results of the affected population and the actual situation is only 1.2%.The accuracy is very high,which is basically consistent with the actual situation,it’s a slow recovery.“No.1 flood in Xijiang in2020”assessment of the affected population and the official population relative error of20.6%,a higher accuracy,roughly in line with the actual,rainfall distribution is more heavily affected areas.The results show that it is feasible to analyze natural disasters such as typhoon and flood based on nightlight remote sensing images,and the macroscopic dynamic monitoring of disaster situation can be realized.(3)Construction of disaster visualization system.Based on Arc GIS Online cloud platform,a disaster visualization system is constructed based on the rapid mapping of disaster data to realize the display of disaster data visualization and the sharing and release of disaster data,to provide governments and the public with intuitive disaster information display,emergency management for natural disasters to provide a platform for collaboration. |
PDF Full Text Request