Construction Application Of Forest Health Evaluation Index System In Protected Areas

Forests are the most important terrestrial ecosystems,which are widely distributed and rich in resources.Its health affects human development and the sustainability of natural resources.The assessment of forest health is the basis for understanding the status quo of forest management and protection and formulating management and protection countermeasures.Because the focus of forest health at different scales is different,and there is no unified standard for the health evaluation index system.Therefore,based on the forest structure and type characteristics,this paper establishes an evaluation index system of two scales of landscape and small class,and conducts multi-scale evaluation of forest health in Shennongjia National Park.Using the fifth forest resources census in Hubei Province and remote sensing data as the basic data,an index system was constructed from two scales of landscape and small class,and the health index of landscape and small class was calculated to evaluate the forest health of the two scales,and finally the random forest algorithm was used to evaluate the results.Construction of predictive models.The main results are as follows:(1)Shennongjia National Park is divided into 25 forest landscape types,and the results of the division can be used as the basic unit of forest landscape health assessment.(2)The forest health evaluation index system was constructed from the two scales of landscape and small class.There are 11 indicators in total at the landscape scale.The index system focuses on 4 aspects of forest landscape composition,shape,diversity and spatial distribution;the small class scale has a total of 14 indicators.The index system can reflect the health status of forests in four aspects: structure,vitality,sustainability and resistance.The health evaluation index system of the two scales reflects the forest health status from different angles and can be used for forest health evaluation.(3)The results of landscape evaluation show that among the 6 types of coniferous forests,2 types are of high-quality health level,3 types are of good health level,and 1 type is of general health level;among the 10 types of broad-leaved forests,2 types are of high-quality health level,and 5 types are of good health level.3 types of health level are general health level;among the6 types of mixed forests,1 type is high-quality health level,2 type is good health level,1 type is general health level,1 type is sub-health level,and 1 type is unhealthy level;2 Shrub-like forests were all at unhealthy levels;bamboo forests were at high-quality healthy levels.In the three main forest types,the health levels of all levels are distributed;the comparison results of forest types of different origins show that there is no significant difference in landscape health status between natural forests and artificial forests.In general,there are few forest landscapes at sub-health and unhealthy levels,and the overall forest landscape of Shennongjia National Park is at a healthy level.(4)Among the 7,242 forest classes in Shennongjia National Park,the number of forest classes with high-quality health grade is the least,only 65,accounting for 0.90% of the total number of classes;there are 1,075 classes with good health level,accounting for 14.84% of the total number of classes.;The number of small classes in the general health level is the largest,with a total of 3662,accounting for 50.57% of the total number of small classes;the number of small classes in the sub-health level is second only to the general healthy class,accounting for29.00% of the total number of small classes;the number of small classes in the unhealthy level is less,There are 340 in total,accounting for 4.69% of the total number of small classes.Overall,the forest classes in Shennongjia National Park are generally at a healthy level,with 66.31% of the forest classes above average health,accounting for about 2/3 of the total,with continuity in spatial distribution.(5)Based on the evaluation results of small classes,two kinds of forest health evaluation models were constructed by using random forest algorithm.The model with simplified features also has good accuracy,so indicators such as average diameter at breast height,soil thickness,unit volume,and LST annual maximum value can be used to quickly evaluate forest health.In this paper,landscape and small classes are selected as the two scales to study the forest health of Shennongjia National Park.On the basis of the forest landscape zoning and the analysis of the characteristics of small classes,the corresponding index system is established,and the forest health status in the study area is evaluated by calculating the forest health index..The application of multi-source data effectively improves the representativeness and reliability of the data.The multi-scale evaluation results more comprehensively reflect the forest health status,and can provide a basis for decision-making of forest ecological protection.The health evaluation model constructed based on the random forest algorithm still has good accuracy under the condition of simplified features,which provides a reference for determining the main indicators that should be paid attention to in the rapid evaluation of forest health.