Spread Risk And Warning Of Hyphantria Cunea(Drury) Based On Nighttime Light Remote Sensing Index

Hyphantria cunea is a major invasive biological quarantine object jointly managed by China’s agricultural and forestry departments.Since its invasion in China in 1979,its occurrence area has increased year by year,and the degree of damage has been increasing,which has seriously affected China’s ecological security.Based on GIS(Geographic information system)raster calculation,the nighttime remote sensing index was introduced to represent the impacts of human on the spread of pests.It is of great significance to carry out scientific spread risk analysis and spatial pattern research for the integrated control practice and protection of agricultural and forestry ecosystems.Despite existing literature records that human activities are the core variable driving the spread of Hyphantria cunea,systematic analysis of the historical and statistical data is insufficient to describe the human dominant propagation.Due to the complexity and diversity of human activities and the difficulty of spatial expression,the existing risk assessment lacks the consideration of this crucial variable,which affects the accuracy of forecast and warning.Relying on interdisciplinary integration,developing new methods of diffusion characteristics and pattern analysis,improving the accuracy of human activities and other variables,and improving the accuracy of warning of spread risk are urgent issues to be solved in the spread warning analysis of Hyphantria cunea.Based on the special investigation of Hyphantria cunea,announcement of epidemic area from the National Forestry and Grassland Bureau Forest Pest Control Station,nighttime light remote sensing data NPP/VIIRS and MODIS etc.,the spatial analysis method of spatial autocorrelation and standard deviation ellipse fitting sample data was proposed to reveal the spatial dispersed characteristic and patterns by integrating GIS(Geographic Information System)geostatistical method theory.The results and conclusions are as follows:(1)The results show that the spread of Hyphantria cunea is random in general,but it has the characteristics of regional aggregation in some parts.In recent years,the annual average dispersed surface velocity has reached 58769km2/a,the linear velocity has reached 131.82km/a,and the diffusion speed in the main direction is 75.21km/a,which has broken through the current statistical conclusion that the annual outward spread of Hyphantria cunea is 35-50 km.Human activities are the dominant factor of the spread.(2)Compared with the human activity variables based on the basic geographic information simulation,the light remote sensing index show better effect on the impact indicator;the overall spread risk accuracy of Hyphantria cunea increased by 3.3% with the introduction of the light remote sensing index.(3)The ROC curve of spread risk warning model shows that the light remote sensing index is the most sensitive and significant variable and human activity is the main influencing variable of diffusion propagation.(4)The early warning results show that in the future,the high-risk and extremely high-risk areas of Hyphantria cunea will be mainly concentrated in the economically developed areas in the East,and will spread southward and westward.The southern region dominated by the middle and lower reaches of the Yangtze River is the main prevention and control area at present.In the long run,Sichuan Basin,the Loess Plateau,the agricultural areas of the Mongolian and Xinjiang plateau,and the urban areas will also be the major areas of diffusion region.(5)The composition of the epidemic areas and the national land use landscape shows that the Hyphantria cunea has formed the spatial pattern of dispersing,colonizing,breeding,developing and harming in the high density area of human activities,taking agricultural vegetation and other industrial vegetation as its host,and spreading through narrow corridors and throat location rather than establishment in mountain vegetation ecosystem.Priority should be given to the monitoring,prevention and control of the spread corridor,throat area and sprout position,which can effectively stop the spread and expansion channel of Hyphantria cunea.Strengthening the restoration and construction of near natural forest ecosystem,maintaining biodiversity and forest health,and building ecological barriers can prevent the trend of expanding to the South,West and North of Hyphantria cunea.(6)Based on GIS spatial matrix model,forecast variables and warning results are presented in 250 m scale spatial unit.The spatial continuous description of risk value is realized in 0~1 probability.The accuracy of warning is improved by visual map expression and interactive understanding,which lays the technical foundation of spatiotemporal forecast and warning,and has obvious spatial indication for quarantine and prevention at local bureau.In this dissertation,we proposed and described quantitatively the spatial analysis method of using sample data to characterize the spread pattern of Hyphantria cunea for the first time.Then a light remote sensing index based on nighttime light remote sensing data,land use and geographical environment impedance was developed to depict human impacts field,and the index was introduced into the spread risk and damage degree prediction of Hyphantria cunea warning,which plays an important role in improving the accuracy of forecasting and warning and establishing the analysis of spatiotemporal forecast.In addition,using 0~1 probability measure and visualized map to express the forecast risk has an obvious indicative role for local bureau to better understand and apply.The spatial analysis method and warning technology presented in this paper have certain reference significance and application value for the risk assessment of other pests and the relationship between human and natrual ecosystem.