Study On Potential Habitat And Occurrence Prediction Of Hyphantria Cunea

Hyphantria cunea Drury is one of the most serious invasive pests in China.At present,the epidemic has spread to most of the central and eastern parts of China,which has a huge impact on the forest ecological resources and economic development of the local forest regions.This study on the spatial distribution of the moths,migration,extent of area,happen,algebra is forecasted and analyzed,and the design and implementation H.cunea forecasting warning system,to develop reasonable H.cunea control countermeasures and key monitoring area to provide theoretical basis and technological means,this experiment has achieved the following results:(1)ROC curve analysis showed that the average AUC values predicted by the random forest model for the potential habitat distribution of H.cunea were 0.98,respectively.Under the current period,the potential habitat of H.cunea accounts for 8.74%of China's land area,and the suitable areas are mainly concentrated in the south of northeast China,the southeast of north China,the north of south-central China and the north of east China.The environmental factors that have great influence on the potential habitat distribution of H.cunea are:altitude,vegetation coverage,average temperature in the wettest season,and average temperature in the warmest season.Under the climate scenario of RCP2.6 in the 2050s,the total living area of H.cunea accounts for 14.38%of China's total land area,and under the climate scenario of RCP8.5,the total living area of the H.cunea accounts for 19.06%of China's land area.In the 2050s,the potential habitat center of the H.cunea will migrate 101.33km to the north on average,and the newly added suitable areas include Heilongjiang province,Jilin province,Sichuan province,Hubei province,Shanxi Province,Henan province,eastern Inner Mongolia and Taiwan island.(2)Based on the random forest model,the occurrence degree of the H.cunea in north China plain is predicted in the short and medium term.The model can predict the occurrence degree of the H.cunea in the next month according to the meteorological data in the first three months of the occurrence period.The overall training accuracy of the model is 89.9%.Based on the law of effective accumulated temperature for predict the occurrence of the H.cunea generation,generation numerical results show that in our country H.cunea range from 0-6.1 generation,the vast majority of areas in China can meet the development of generation 1 moths of temperature conditions,most of the southern region climate conditions can make H.cunea happened more than 4 generations.(3)Based on WebGIS technology,the use of a series of open source components to H.cunea forecasting early warning system to study,the system realized with the happening of the H.cunea data query,statistics,and other functions,has realized the real-time processing on the basis of the principle of flow is used to calculate storage meteorological data,build the model base system,according to the forecast model to simulate the H.cunea related prediction research and display the results.