基于150株样木的生物量数据,采用常规回归方法和哑变量模型方法,构建了黑龙江、吉林两省云杉地上总生物量与各分项生物量的一元、二元以及三元模型。结果表明:2种方法建立的模型中,地上总生物量模型预估精度最高,在96%以上;树叶生物量最低,仍达87%以上;其他生物量预估精度均在91%以上;总相对误差均控制在±5%的范围内。通过对一元、二元和三元模型对比分析,发现随着解释变量增加,2种方法生物量模型的预估精度和确定系数都有所提高。引入地域哑变量后,一元、二元和三元模型的预估精度和确定系数都比常规模型有所提高,估计值的标准误差和总相对误差有一定下降,哑变量可以提高模型的拟合优度和预测效果。
Based on the biomass data of 150 spruce sampling trees,by using conventional regression methods and dummy variable modeling approach,one variable and two or three variables biomass models were established for the total aboveground biomass and the biomass of components for spruce in Heilongjiang and Jilin provinces.The results showed that the total aboveground biomass models had the highest prediction accuracy(96% or more)and the leaf biomass models had the lowest prediction accuracy which still reached more than 87%.The prediction of other models reached more than 91% and the total relative error was controlled within ±5%.The prediction accuracy and the determination coefficient of biomass models were improved with the increase of the explanatory variables.After the introduction of the dummy variable,the prediction accuracy and the determination coefficient was improved meanwhile the standard error and the total relative error was reduced for one variable and two or three variables biomass models.Dummy variable can improve the prediction effect of the model.
[1]TomppoE,GschwantnerT,LawrenceM,et al.National Forest Inventories:Pathways for Common Reporting [M].NewYork:Springer,2010.
[2]王鑫.北京地区刺槐地上部分生物量模型研究[D].北京:北京林业大学,2011.
[3]唐守正,张会儒,胥辉.相容性生物量模型的建立及其估计方法研究[J].林业科学,2000,36(1):19-27.
[4]曾伟生.全国立木生物量方程建模方法研究[D].北京:中国林业科学研究院,2011.
[5]张连金,曾伟生,唐守正.用带截距的非线性方程和分段建模方法对立木生物量估计的比较[J].林业科学研究,2011,24(4):453-457.
[6]符利勇,曾伟生,唐守正.利用混合模型分析地域对国内马尾松生物量的影响[J].生态学报,2011,31(19):5797-5808.
[7]李希菲,洪玲霞.用哑变量法求算立地指数曲线族的研究[J].
林业科学研究,1997,10(2):215-219.
[8]李河,麦劲壮,肖敏,等.哑变量在Logistic回归模型中的应用[J].循证医学,2008,8(1):42-45.
[9]郑冬梅,曾伟生.用哑变量方法构建东北落叶松和栎类分段地上生物量模型[J].北京林业大学学报,2013,35(6):23-27.
[10]曾伟生,唐守正,夏忠胜,等.利用线性混合模型和哑变量模型方法建立贵州省通用性生物量方程[J].林业科学研究,2011,24(3):285-291.
[11]国家林业局.LY/T 2258-2104,立木生物量建模方法技术规程[S].北京:中国标准出版社,2014.
[12]Jenkins J C,Chojnacky D C,Heath L S,et al.National-scale biomassestimators for United States treespecies[J].For.Sci.,2003,49(1):12-35.
[13]Muukkonen P.Generalized allometric volume and biomass equations for some tree species in Europe[J].Eur.J.Forest Res.,2007,126:157-166.
[14]N·varJ.Allometric equations for tree species and carbon stocks for forests of northwestern Mexico[J].Forest Ecology and Management,2009,257:427-434.
[15]Parresol B R.Additivity of nonlinear biomass equations[J].Canadian Journal of Forest Research,2001,31(5):865-878.
[16]Zhang Y J,Borders B E.Using a system mixed-effects modeling method to estimate tree compartment biomass for intensively managed loblolly pines-an allometric approach[J].Forest Ecology and Management,2004,194(1/3):145-157.
[17]Bi H,Turner J,Lambert MJ.Additive biomass equations for native eucalypt forest trees of temperate Australia[J].Trees,2004,18(4):467-479.
[18]Fehrmann L,Lehtonen A,KleinnC,et al.Comparison of linear and mixed-effect regression models and a k-nearest neighbour approach for estimation of single-tree biomass[J].Canadian Journal of Forest Research,2008,38(1):1-9.
[19]胥辉.立木生物量模型构建及估计方法的研究[D].北京:北京林业大学,1998.
[20]张会儒,唐守正,王奉瑜.与材积兼容的生物量模型的建立及其估计方法研究[J].林业科学研究,1999,12(1):53-59.
[21]高东启,邓华锋,蒋益,等.油松林分断面积生长预估模型研究[J].西南林业大学学报,2015,35(1):42-46.
[22]张会儒,唐守正,胥辉.关于生物量模型中的异方差问题[J].林业资源管理,1999(1):46-49.
[23]曾伟生,骆期邦,贺东北.论加权回归与建模[J].林业科学,1999,35(5):5-11.
[24]曾伟生.再论加权最小二乘法中权函数的选择[J].中南林业调查规划,1998,17(3):9-11.
