林木虫害大数据的网络科学分析方法

摘要 择取国家森防总站2009—2013年辽宁省林木虫害大数据，根据林木虫害时空复杂性，提出一种基于时空影响域的虫害关系网络构造方法.以昆虫生活习性确定时间影响范围，以虫害危害等级确定空间影响范围，并将松毛虫作为研究对象.结果表明，松毛虫虫害关系网络为无标度网络，服从幂律分布；松毛虫传播扩散快；虫害易聚集发生；网络拓扑具有鲁棒性.该建网方法能够反映真实世界，是解读林木虫害大数据的有效方法.期待通过本文对松毛虫虫害关系网络的复杂网络理论分析，能对实际林木虫害防治工作中防治策略的制定及防控力度的估计提供理论指导.
Abstract：Big data of Liaoning forestry pests was adopted， which publicated by General Station of Forest Pest Management， State Forestry Administration. A construction method based on space-time influence domain for insects network was proposed according to the insects occurring complexity in space and time. Taking pine caterpillar (Dendrolimus spp.) as a research sample， the expansion of the time window is determined according to the insects’ lifestyles and habits， and the range of influence effect are determined according to the damage of pest. The results show that the proposed pine caterpillar network following the power-law distribution is scale-free， Dendrolimus diffuse fast， the pine caterpillars are more likely to emerge clustered， and the topology of the pine caterpillar network is robust. This analysis which using the complex network method is scientific， and the real-world phenomenon can be reflected in the network construction method. Such a analysis of network science on pine caterpillar network is intended to provide a guidance regarding to forestry pest control strategies.