科尔沁沙地2种优势固沙灌木的相容性生物量模型

Compatible Biomass Models for Two Dominant Sand-fixing Shrubs in Horqin Sandy Land

基于现有的灌木生物量模型存在各组分与总量不相容问题,以半干旱区科尔沁沙地2种优势固沙灌木小叶锦鸡儿(Caragana microphylla)和黄柳(Salix gordejevii)为对象,通过对大样本的相关形态指标及生物量实测数据进行非线性回归和非线性误差变量联立方程组的计算,建立了2种灌木叶片、新枝、老枝和地上生物量的独立模型及与各组分相容的生物量模型,并通过决定系数R^2\,估计值的标准差SEE\,总相对误差TRE\,平均系统误差MSE(ASE)\,平均预估误差MPE和平均百分标准误差MPSE对模型进行评价。结果表明:非线性加权回归使得两种灌木各组分的独立模型拟合效果优化,而生物量相容模型与独立模型相比,由于抽样误差的存在拟合效果不佳,相关评价指标较低。

Based on the existing shrub biomass model,there is a problem that the components are incompatible with the total amount.In this study,the independent model of new shoot,old shoot,aboveground biomass and the model which was compatible with each component were constructed after the nonlinear regression and nonlinear error variables simultaneous equations were calculated by comparing the relevant morphological indicators and biomass measured data of large samples about two dominant sand-fixing shrubs,Caragana microphylla and Salix gordejevii,were studied in Horqin Sandy Land.The model was evaluated by the largest determination coefficients(R^2),the smallest standard error of estimates(SEE),the total relative error(TRE),the average systematic error(MSE/ASE),the mean percent estimate error(MPE)and the average percentage standard error(MPSE).The results showed that the nonlinear weighted regression optimized the independent model fitting effect of the two shrubs,compared with the independent model,the biomass compatible model had poor fitting effect due to the existence of sampling error,and the relevant evaluation index was low.