应用结构关系识别法对黑龙江省森林碳储量的预测  

Forecast of Forest Carbon Stock in Heilongjiang Province with Structural Relationship Identification

作　　者：王姝雅 高丹丹[1] 刘鹏[1] 李顺龙[1] Wang Shuya;Gao Dandan;Liu Peng;Li Shunlong(Northeast Forestry University,Harbin 150040,P.R.China)

机构地区：[1]东北林业大学,哈尔滨150040

出　　处：《东北林业大学学报》2018年第9期59-64,共6页Journal of Northeast Forestry University

摘　　要：采用森林蓄积量扩展法,计算黑龙江省1986—2016年森林碳储量。从林业碳保护管理行为、碳吸引管理行为、林业经济发展规模三个方面,选取森林病虫害发生面积、森林火灾受害面积、造林面积、森林抚育面积、森林采伐量、林业基本建设投资、林业总产值7个指标,结合黑龙江省1986—2016年间森林碳储量相关指标数据,运用格兰杰因果检验、弱外生性检验识别森林碳储量与变量间的联系;通过Johansen协整检验,分析内生变量是否存在长期均衡关系;选用自回归滑动平均(ARMA)模型、向量自回归(VAR)模型、向量误差修正(VECM)模型进行样本内预测,比较模型预测准确性。结果表明:森林病虫害发生面积、林业总产值,与森林碳储量间存在格兰杰因果关系;造林面积通过影响林业总产值,间接影响森林碳储量;森林火灾受害面积与森林病虫害发生面积存在关联;可采用预测结果最优的VECM模型对黑龙江省森林碳储量进行样本外预测。The forest stocks expansion method was used to calculate the forest carbon stocks in Heilongjiang Province from 1986 to 2016.From forestry carbon protection management behavior,carbon attraction management behavior and forestry economic development scale,seven indicators were selected including forest pest and disease occurrence area,forest fire damage area,afforestation area,forest rearing area,forest harvesting amount,forestry capital construction investment,and total forestry output value.Combined with data on forest carbon stocks in Heilongjiang Province from 1986 to 2016,Granger causality test and weak exogenous were used to identify the relationship between forest carbon storage and variables.The performance was analyzed that whether there was a long-term equilibrium relationship between endogenous variables through the Johansen cointegration test.In-sample prediction was performed using ARMA model,VAR model and VECM model,and the prediction accuracy of the model was compared.There is Granger causality between the occurrence area of forest pests and diseases,total output value of forestry and forest carbon storage.Afforestation area indirectly affects forest carbon storage by affecting the total output value of forestry.The area affected by forest fire is related to the occurrence area of forest pests.The VECM model with the best prediction results can be used for out of sample prediction of forest carbon storage in Heilongjiang Province.

关 键 词：森林碳储量 结构关系识别 向量误差修正模型 黑龙江省 

分 类 号：S718.5[农业科学—林学]