Methods Of Forest Health Assessment Based On Forest Inventory And Forest Management Inventory Data

Forest health assessment is an emerging field of forestry research.Scientific and rational forest health assessment is significant for protecting forest resources and improving forest management effectively.Chun'an county was taken as the research area in this paper.The checked continuous forest inventory and forest management inventory data in 2017 were the main data sources.The forest health evaluation index system for trees was constructed firstly,using the relevant indicators of the checked continuous forest inventory data,and then the forest health theory value is obtained.The neural network model is constructed with the corresponding forest management inventory data to get the more accurate predictions,and a better forest health evaluation model based on generalized neural network is finally completed.The main results of the paper are as follows:(1)On the basis of reviewing the literature on forest health evaluation,404 sample data of trees were selected in the 735 continuous forest inventory data.From the four levels of productivity,organizational structure,site conditions and management,22 indicators with strong correlation were selected as forest health evaluation indicators.The forest health evaluation index system was established and the weight values of each index were calculated using Analytic Hierarchy Process(AHP).(2)Using the established forest health evaluation index system,the forest health index and forest health level of the 404 sample plots of trees were calculated.The forest health index range is 0.46-0.84;47 sample plots were health,accounting for 11.63%;207 sample plots were sub-health,accounting for 51.24%;130 sample plots were medium health,accounting for 32.18%;20 sample plots were unhealth,accounting for 4.95%;The relationship between the continuous forest inventory and forest management inventory data was established firstly,and then the forest health assessment values of the continuous forest inventory plot were assigned to the corresponding forest resources inventory data.(3)Using the generalized regression neural network method,based on the forest resources inventory data given the theoretical value of forest health.Two models with 18 indicators and 10 easier-to-obtain indicators were separately establish for learning training and predictive verification experiment.324 of the 404 forest resources inventory data were used as learning training samples,and the remaining 80 samples were used as verification data.The results show that the average relative error of model with 10 indicators is 0.0710,while the model with 18 indicators is 0.0721.The forest health assessment model with 10 easier-to-obtain indicators based on generalized regression neural network can get relatively accurate predictions.