基于SOM神经网络的宜城市森林健康评价  

Forest health assessment of Yicheng City based on SOM neural network

作　　者：林治成 杨贵才 夏锐 康建坤 曾冠中 张伟[2] Lin Zhicheng;Yang Guicai;Xia Rui;Kang Jiankun;Zeng Guanzhong;Zhang Wei(School of Geophysics and Geomatics,China University of Geosciences(Wuhan),Wuhan Hubei 430074,China;Center for Geophysical Survey,China Geological Survey,Langfang Hebei 065000,China;Natural Resources Comprehensive Survey Command Center of China Geological Survey,Beijing 100055,China)

机构地区：[1]中国地质大学(武汉)·地球物理与空间信息学院,湖北武汉430074 [2]中国地质调查局地球物理调查中心,河北廊坊065000 [3]中国地质调查局自然资源综合调查指挥中心,北京100055

出　　处：《南方林业科学》2023年第5期65-70,共6页South China Forestry Science

基　　金：中国地质调查局项目(项目编号:ZD20220130)。

摘　　要：森林健康是生态文明建设中不可或缺的一环。本文基于2022年宜城市森林调查数据,通过主成分分析法排除冗余因子,确定健康评价指标,并运用SOM神经网络模型进行聚类分析,对宜城市森林健康状况进行评价。结果表明:(1)总的来看,宜城市森林优质健康等级最少,占比12.86%,健康等级占比17.14%,亚健康等级占比21.43%,不健康等级最多,占比31.43%,极不健康等级占比17.14%;(2)宜城市森林以幼龄林与中龄林为主,龄组结构较为不合理,中龄林健康状况整体略优于幼龄林;(3)以郁闭度划分,低郁闭度森林健康状况优于高郁闭度森林;(4)以起源划分,人工林健康状况优于天然林。Forest health is an indispensable part of the construction of ecological civilization.Based on the data of forest survey in Yicheng City in 2022,redundant factors were excluded by principal component analysis method to determine the health evaluation index,and SOM neural network model was used for cluster analysis to evaluate forest health status in Yicheng City.The results showed as follows:(1)In general,the forest of Yicheng had the least quality health grade(12.86%),the health grade(17.14%),the sub-health grade(21.43%),the unhealthy grade(31.43%)and the extremely unhealthy grade(17.14%).(2)The forest in Yicheng city was mainly composed of young forest and middle-aged forest,the age group structure was not reasonable,and the health status of middle-aged forest was slightly better than that of young forest;(3)The health status of forest with low canopy closure was better than that of forest with high canopy closure;(4)According to the origin,the health status of planted forest was better than that of natural forest.

关 键 词：森林健康评价 宜城市 主成分分析法 SOM神经网络 聚类分析 

分 类 号：S718.5[农业科学—林学]