基于SOM神经网络的新化县森林健康评价

Forest health assessment of Xinhua county based on SOM neural network

运用主成分分析法、定量和定性相结合的方法构建了新化县森林健康评价指标体系,基于SOM神经网络构建了新化县森林健康评价模型,对新化县森林进行健康评价,利用GIS技术对健康评价结果和森林属性进行了数据对比分析,结果表明:新化县森林的总体健康水平不高,主要以较健康状态为主;森林的健康状况存在明显的空间分布特征,森林的健康状况与人为干扰程度存在明显的相关关系;结合森林起源、林种、龄组、主要优势树种和森林健康等级空间分布进行了对比分析,得出了新化县森林起源、林种、龄组和主要优势树种健康状况的现状和成因。

In order to build the Xinhua county forest health evaluation system using principal component analysis, quantitative and qualitative methods, map neural network constructed Xinhua county forest health assessment model based on SOM self-organizing feature of Xinhua county forests health assessment. Health assessment and forest attributes were analyzed by GIS technology and the results show that the overall health of Xinhua county forest is not high, mainly in the relatively healthy state based high-quality health; the health status of forest has obvious spatial distribution characteristics, which existing significant correlation with human disturbance. Comparative analysis was carried out combining with origin of forest, category, age group, and dominant tree species with spatial distribution of forest health. Thus, the status quo and the causes about category of forest origin in Xinhua county was concluded.