Remote Sensing Monitoring And Predicting For Pristiphora Laricis(Hartig) Of Saihanba Mechanical Forestry Farm

Forest pests are considered as one of the typical biological and natural disasters, the infestation of which was more and more serious. Remote sensing, GIS (Geographic Information System) and GPS were effective technologies for monitoring and predicting of forest pests, which have important significance for monitoring and precaution, disaster prevention and reduction of forest pests. In this paper, Saihanba Mechanical Forestry Farm of Hebei Province was taken as the study area, and the outbreaks of Pristiphora laricis(Hartig) was taken as the study background. Using the survey data for Pristiphora laricis(Hartig) as well as stand factors and topographic data, the relationship between the occurrence of Pristiphora laricis(Hartig) and habitat factors were analyzed; Using remotely sensed images combined the measured spectral data at the same time, this paper also analyzed the change of the spectral features under different damaged degrees, and built a monitoring model of Pristiphora laricis(Hartig) infestation based on the vegetation index; Then, spatial prediction for Pristiphora laricis(Hartig) infestation in the study area were performed based on GIS and information value model. Supported by this model, a spatial prediction map with 5 classes (extremely high, high, middle, low, and none) was obtained, and the prediction model precision was also analyzed.The main research contents and results in this paper are as follows:1) According to forest resource inventory data for management, the occurrence of Pristiphora laricis(Hartig),meteorological data, ground measured spectral data, GPS data, topographic map, forest resource map, damage degree map, and remote sensing images, A GIS database of study area were built.2) The factors influencing the population of the pest were analyzed. The results indicated that the occurrence of Pristiphora laricis(Hartig) was closely related to the forest stand conditions, topographic and climatic factors. The infestation of Pristiphora laricis(Hartig) was more serious in pure forest, open forest land, mature forest, higher altitude stands, southern slope and ridge than in mixed forest, dense stands, half-mature and young forest, lower altitude stands, northern slope and valley; it was also more serious along the forest boundary than in the forest; the infestation was more serious under the circumstance of small perpendicular height; And the higher was the temperature, the earlier the larva hatched. In addition, the activities of human (controlling measures) can reduce the population density effectively.3) Larix principis-rupprechtii spectrum features at green, red and near infrared bands were studied by measuring and analyzing its canopy reflectance and differential spectrum under different level damaged of Pristiphora laricis(Hartig). The results showed that the spectral reflectance rate at green, red and near infrared bands will respectively descend, ascend and descend with Pristiphora laricis(Hartig)'s outbreak, and the maximal value of first derivative will descend with the damage increased. Moreover, the spectrum moves to the direction of short wave. Another result was that NDVI extracted from spectral data was negatively correlated with damage degrees, which had an important indication significance for the early infection forecast of Pristiphora laricis(Hartig) with remote sensing.4) The relationship between NDVI and the population density was studied, and monitoring model of Pristiphora laricis(Hartig) infestation based on the vegetation index was built as p opulation density= ?1287.878+1946.733/NDVI, which can be used in monitoring Pristiphora laricis(Hartig) based on the vegetation index. The results indicated that a negative correlation existed in the relation between NDVI and the population density. In other words, with the decrease of NDVI, the population density of Pristiphora laricis(Hartig) appears linearly increased. This implicated that NDVI is a good indicators of Pristiphora laricis(Hartig) location, so that can be used as a predictor for Pristiphora laricis(Hartig) remote monitoring.5) Spatial prediction for Pristiphora laricis(Hartig) infestation in the study area was studied based on GIS and information value model.Supported by this model, a spatial prediction map with 5 classes(extremely high, high, middle, low, and none)was obtained. and the precision can be up to 83.1% which indicated that prediction model is very practical. The prediction map can provide scientific references for Pristiphora laricis(Hartig) controlling.