基于非线性联立方程系统的杉木生物量动态预测  

Dynamic model for Cunninghamia lanceolata biomass based on nonlinear simultaneous equations system

作　　者：付尧 尹晶萍 许昊[2,3] 高志雄 马俊杰 FU Yao;YIN Jingping;XU Hao;GAO Zhixiong;MA Junjie(Industry Development and Planning Institute,National Forestry and Grassland Administration,Beijing 100010,China;School of Economics and Management,Ningxia University,Yinchuan 750021,Ningxia,China;School of Forestry and Prataculture,Ningxia University,Yinchuan 750021,Ningxia,China)

机构地区：[1]国家林业和草原局产业发展规划院,北京100010 [2]宁夏大学经济管理学院,宁夏银川750021 [3]宁夏大学林业与草业学院,宁夏银川750021

出　　处：《中南林业科技大学学报》2025年第4期151-161,共11页Journal of Central South University of Forestry & Technology

基　　金：福建省林业科学技术研究项目(2023FJLY0504101013)。

摘　　要：【目的】森林碳汇是实现“碳中和”战略目标的重要路径之一。准确评估森林生物量的动态变化,对掌握森林碳汇功能,支撑区域战略落地和区域目标实现有着重要的作用。【方法】以福建将乐的杉木Cunninghamia lanceolata为研究对象,通过样地调查、样木采集、数据整理与分析等方法,利用R语言,基于Logistics生长方程和幂函数结构建立杉木胸径、树高与年龄的理论生长方程和杉木地上、地下以及总生物量与胸径、树高的异速生长方程。在此基础上,利用非线性最小二乘法(NOLS)和非线性似不相关回归(NSUR),构建杉木地上生物量、地下生物量以及总生物量的非线性联立方程系统,并对其进行动态预测研究。【结果】1)基于非线性联立方程系统建立的杉木地上、地下以及总生物量的动态变化模型相较于传统杉木生物量生长模型具有较高的拟合精度,其中,基于NSUR建立的模型较传统模型的平均绝对误差(AME)、剩余均方根误差(RMSE)分别下降30.99%和33.68%,而R^(2)提升了7.78%;2)不同方法的预估能力不同,利用NSUR方法建立的模型相较于NOLS方法的拟合精度和验证精度均更高;3)杉木胸径生长对杉木生物量的异速生长呈现正向影响(协方差值>0),而树高生长对杉木生物量呈现负向影响(协方差值<0)。【结论】基于杉木生长因子的理论生长方程与生物量异速生长方程的非线性联立方程系统能够更好地反映杉木生长与生物量间的复杂关系,本研究建立的模型对开展杉木生物量的动态预测、制定科学合理的经营管理措施具有重要的指导意义。【Objective】Forest carbon sequestration is a crucial pathway for achieving the strategic goal of“carbon neutrality”.Accurately assessing the dynamic changes in forest biomass is important for understanding forest carbon sequestration functions and supporting the implementation of regional strategies and goals.【Method】This study focused on the Cunninghamia lanceolata species in Jiangle,Fujian Province.Through plot surveys,sample tree collection,data sorting,and analysis,theoretical growth equations for the diameter at breast height(D),tree height(H),and age of Chinese firs were established based on the logistics growth equation and power function structure,using R language.Additionally,allometric growth equations were developed for aboveground biomass(AGB),underground biomass(UGB),and total biomass(TB)in relation to D and H.On this basis,nonlinear least squares(NOLS)and seemingly unrelated regression(NSUR)were used to construct a nonlinear simultaneous equation system for predicting the dynamic changes in aboveground biomass,belowground biomass,and total biomass of Chinese firs.【Result】1)The dynamic change model for aboveground,belowground,and total biomass of Chinese fir established based on the nonlinear simultaneous equation system has higher fitting accuracy compared to traditional biomass growth models.Specifically,the models based on NSUR reduced AME and RMSE by 30.99%and 33.68%,respectively,while R^(2) increased by 7.78%;2)Different methods have different predictive abilities,with the models established using the NSUR method showing higher fitting accuracy and validation precision compared to those using the NOLS method;3)The D growth of Chinese firs positively influences the allometric growth of biomass(covariance value>0),whereas the H growth negatively influences the biomass(covariance value<0).【Conclusion】The nonlinear simultaneous equation system based on theoretical growth equations and allometric growth equations of Chinese fir factors can better reflect the complex relationships b

关 键 词：树木特征因子 理论生长方程 异速生长方程 非线性联立方程系统组 杉木 

分 类 号：S791.27[农业科学—林木遗传育种]