湖南省阔叶树资源数据融合与模型研究

发布时间：2018-06-08 23:37

本文选题：林分生长量模型 + 阔叶树资源数据融合　；参考：《中南林业科技大学》2014年硕士论文

【摘要】：森林资源数据是一种随年度而不断变化的数据。森林资源数据主要是通过森林资源连续清查简称一类清查和森林资源规划设计调查简称二类调查获得,由于两套数据调查间隔期、抽样方法和精度要求不同,导致同一地区存在两套森林资源总量数据。近年来,国家林业局森林资源管理司提出,要积极推进森林资源监测体系优化改革,逐步建成服务高效的森林资源一体化监测体系,最终实现国家和地方森林资源监测工作“一盘棋”,森林资源“一套数”,森林分布“一张图”。森林资源一体化监测的核心就是要国家一类清查和地方二类调查数据有效融合。森林资源数据融合的关键在于充分利用两类数据的优点,准确选择合适的本底和适合的模型。本研究在充分总结分析前人对森林资源数据融合模型建模和林分生长模型的基础上,以湖南省主要阔叶树,包括阔叶纯林、阔叶混交林和针阔混交林为研究对象,利用湖南省森林资源连续清查从1994年到2009年四期样木数据和2004与2009年阔叶树样地数据,利用林分生长模型,构建湖南省主要阔叶树直径、树高及每公顷株数变化模型。主要研究内容和结论如下： (1)把湖南省全省按照不同阔叶树生长所需气候及树种自然分布情况,划分为湘北、湘中东滨湖盆地丘陵区,湘南山地丘陵区和湘西山区三个分区。 (2)各分区按照不同阔叶树种组,分别用六个经典的生长量模型对阔叶树树高生长进行拟合,通过对比各方程拟合的决定系数R2及剩余均方差MSE,确定了最适合该分区阔叶树种组的树高生长最优模型,对树高生长模型求时间的一阶导数,得到了各分区不同阔叶树种组的树高连年生长量模型。 (3)对于各分区阔叶纯林通过计算各变量与平均直径的皮尔森简单相关系数及多变量多重共线性检验,选择了影响直径生长的主要因变量：平均年龄、平均树高、林分密度指数,用多元线性回归模型拟合得出各分区阔叶纯林直径生长模型。对直径生长模型进行△D=Dt+1-Dt运算,得到各分区不同阔叶树纯林的直径连年生长量模型。 (4)各分区针阔混交林通过采用隐去年龄的生长模型法对平均直径进行建模,得到了该分区针阔混交林平均直径变化模型。 (5)利用2004年所选各分区样地相对应的2004年和2009年的样木数据,分别建立其保留木株数模型和进界木株数模型,从而得出每公顷株数模型。 (6)以湖南省江华县2007年的森林资源小班数据为例,采用本研究所得该分区模型系,实现了江华县阔叶树结果数据和连清结果数据在2013年时的融合。
[Abstract]:Forest resource data is a kind of data that varies with the year. The forest resources data are mainly obtained by two types of investigation, namely, the continuous inventory of forest resources and the survey of forest resources planning and design. Because of the two sets of data survey intervals, the sampling methods and the accuracy requirements are different, resulting in the existence of two sets of forests in the same area. In recent years, the Department of forest resources management of the State Forestry Bureau has proposed to actively promote the optimization and reform of the monitoring system of forest resources, gradually build up an integrated monitoring system for the efficient and efficient forest resources, and finally realize the "one game" of national and local forest resources monitoring, "one set of forest resources", and the forest distribution "1." The core of the forest resource integration monitoring is to integrate the national inventory and the local two types of survey data effectively. The key of the forest resource data fusion is to make full use of the advantages of the two types of data and to select the right base and suitable model.
On the basis of summarizing and analyzing the previous model of forest resource data fusion model and forest growth model, the main broad-leaved tree in Hunan Province, including broad-leaved pure forest, broad leaved mixed forest and coniferous and broad-leaved mixed forest, was used as the research object. The data of four period wood samples from 1994 to 2009 were checked continuously from the forest resources in Hunan province and 2004 and 2009. The main research contents and conclusions are as follows: the major broadleaved tree diameter, tree height and plant number per hectare change model in Hunan province.
(1) according to the climate and natural distribution of different broad-leaved trees in Hunan Province, it is divided into three regions in the north of Hunan, the hilly area of the Binhu basin in the Middle East of Hunan, the hilly region of southern Hunan and the mountain area in Xiangxi.
(2) according to the different broad-leaved tree species, six classical growth models were used to fit the high growth of broad-leaved tree tree respectively. By comparing the determination coefficient R2 and the residual mean variance MSE of the equation fitting, the best growth model of tree height suitable for the broad leaved tree species group was determined, and the first derivative of the time was calculated for the tree height growth model. The tree height increment models of different broad-leaved tree species in different districts were obtained.
(3) for the Pearson simple correlation coefficient and multivariable multi collinearity test of the various zoning broad-leaved pure forests, the main dependent variables affecting the diameter growth were selected: average age, average tree height, stand density index, and multivariate linear regression model was used to fit the diameter growth model of the zoning broad-leaved pure forest. Based on the diameter D=Dt+1-Dt growth model, the diameter increment model of different broad-leaved trees in different districts is obtained.
(4) the average diameter was modeled by the growth model of hidden age, and the average diameter change model of the coniferous and broad-leaved mixed forests in the zoning was obtained.
(5) using the data of the sample plots of 2004 and 2009 of the selected subregions selected in 2004, the model of the number of reserved wood number and the number of the incoming wood number are established respectively, thus the number of plant number per hectare is obtained.
(6) taking the data of the small class of forest resources in Jiang Hua County of Hunan province in 2007 as an example, the results of this study were used to achieve the integration of the data of broad-leaved trees in Jiang Hua county and the data of continuous clearing results in 2013.
【学位授予单位】：中南林业科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：S757.2

【参考文献】