杉木幼林地表可燃物含水率对主要火环境因子的响应模型

为探索地表可燃物含水率对主要火环境因子的响应机制,于2003年和2004年的两个春季对福建省南平市原福建林学院后山的杉木Cunninghamia lanceolata人工幼林进行了定点观测,以地表可燃物含水率为因变量(y),以风速(x1)、空气相对湿度(x2)、空气温度(x3)和地表温度(x4)为自变量,采用多项式逼近和多元回归建立了单因子多项式函数、多元线性回归模型,并对模型的复相关系数和偏相关系数进行了t检验和实际验证。结果表明:风速不适于参与模型的建立,其余3个因子参与建立的三元线性模型y=42.345 0.736x2-1.011x3-0.981x4为最佳;模型y=33.406 0.541x2-1.538x3和y=19.049 0.630x2-1.147x4也获得了较高的精度。建议对空气相对湿度和空气温度构成的双因子响应模型进一步研究,以便获得高效快捷的估算地表可燃物含水率的指标。表3参16

Response models on the moisture change of surface fuel to fire environment in Cunninghamia lanceolata young plantation

In the springs of 2003 and 2004,the main fire environment and the moisture content(y) of surface fuel in Cunninghamia lanceolata young plantation at southwest near to Fujian Forestry College,Nanping were continuously inspected and determined in order to illuminate their interrelation.In this paper,the independent variables of wind speed(x_1),air relative humidity(x_2),air temperature(x_3) and the land surface temperature(x_4),and the moisture content of surface fuel(y) are presented with the polynomial approximation.And the multivariate linear models of x_1,x_2,x_3,x_4 are established by the multivariate linear regression.After that,the authors tested the significant of multiple correlation coefficients and partial correlation coefficients with ttest,and calculated also the simulation precision of the selected three models.The results show that wind speed is not proper to construct regression models with the moisture content,on the contrary,the model with x_2,x_3,x_4,is best(y=42.345 0.736x_2-1.011x_3-0.981x_4).y=33.406 0.541x_1-1.538x_3 and y=19.049 0.630x_2-1.147x_4 are also more precision.Lastly,the authors suggest that the response models with x_2 and x_3 should have been more studied in order to find a fast,accurate index to replace the moisture content.