Monitoring And Evaluation The Spatial And Temporal Dynamic Changes Of Corn Armyworm Based On Remote Sensing Data

Heilongjiang is the prime grain producing areas in China as a big agricultural province.The diseases and pests seriously threat to food security, but the traditional methods are difficult to accurately monitor the spatial and temporal dynamic changes of diseases and pests. The disaster area of corn armyworm in Lanxi county of Heilongjiang province was selected as the research object, Hyperspectral and multispectral remote sensing data was used to quantitatively analyze corn canopy and diagnose the insect pest, and determined the optimal monitoring method. To develop dynamic monitoring and evaluation methods of corn armyworm based on multi-temporal remote sensing images through multiple spatial and temporal changes of vegetation index. It reveals the spatial and temporal changes process of corn armyworm. Comprehensive utilization remote technology and weather atmosphere scene, environment information to analysis the change of pests’ weather atmosphere scene and environment feature, which will help monitor pests’ disaster. The results of the research showed that:Satellite hyperspectral data can be used for insect pest monitoring.Based on hyper spectral data to the diagnosis insect pest of maize canopy, Characteristics in visible light blue, green is not suitable for disaster monitoring. In700～760nm red edge position, with its disaster degree aggravating, the slope decreases. First order differential transform data in blue, and green wavelengths pests difference is not obvious, in the702nm to795nm can tell the difference between different levels of disaster, which can be used to select a band edge feature area. Through the analysis of sensitive wave bands, satellite high spectrum can be used for pest monitoring; the sensitive wavelength is660nm to880nm. D-value of red band and infrared wavelengths difference is very sensitive to the plants quantity measurement. The heavier disaster, the smaller difference is.Time series of multispectral data reveals the changing process of insect pests. From multiple period of multispectral data analysis, within the scope of the visible light, there are no blue, red, band absorption and the apparent characteristics of reflection of green light as green vegetation, the near infrared wave band can see highest disaster is different from other disasters; in the late disasters, within the scope of the visible light, the blue, green and red bands, battered reflectivity slants big, obviously at near infrared wave band, with the disaster of reflectivity is small; The heavier the same period, the degree of disasters, the smaller the near infrared band reflectance, the smaller the slope. With the passage of time, the slope becomes smaller, more significant difference between different levels of disaster. The more vegetation damaged, the greater difference of its spectral characteristics is. To monitor the pests in near infrared wave band and red edge slope, multispectral data is more sensitive than the hyper spectral data, the differences between different disasters are more apparent. It can improve the precision of remote sensing monitoring crop condition.Different data sources have different advantages. By constructing three vegetation index analysis the disasters, vegetation index difference of building group is better than high spectrum, hyper spectral group difference is better than multispectral, the same vegetation index has the same features, RDVI and SAVI different disaster level are greater than the NDVI, RDVI variance value is minimum in it; Through building multi-source data vegetation index model, hyper spectral heavy normalized difference vegetation index RDVI monitoring pests clear difference between different level of disaster.Using the spectral reflectance vegetation index model can better build multiple periods from the reaction of the vegetation stress changes. Analysis of five times RDVI difference with different levels of disasters and the change slope, pest infestation, the affected leaves early aging, aging speed varies between different levels of disaster. Early monitoring, the light spectrum similar disasters and health, and not easy to distinguish, presents the plagues of light distribution in healthy and classification results; In the late monitoring, disaster and the battle is not easy to distinguish; In the whole remote sensing monitoring period, which battered monitoring more accurate, as the change of corn leaf physiological parameters, disaster degree different happens, light disaster affected leaf will resume growth over time, most will be aging. Affected by leaf aging speed, different level disaster is suitable for different time. The time between August24and31is most suitable to monitor pest disaster. Pests disaster get about10effectively control monitoring as the best time. Spatial variation characteristics of vegetation index, multiple period of corn stick insect disaster when space change process.Through the analysis of habitat factors, not by coercion or pest vegetation index model of light by pests stress is closely related to the canopy water, early disaster can predict the change of water through the canopy, but for serious intimidation of corn, through correlation analysis, the relationship between temperature and canopy water significantly, and disaster degree was associated with a significant level. It provides a good condition for the growth of armyworm, and lead to armyworm rampant throughout the region. Research results to forecast the possibility development trend of pests, which can improve accuracy of the remote sensing of crop condition monitoring and yield estimation.