论文部分内容阅读
对春季和秋季大兴安岭地区西林吉林业局山杨-白桦混交林、落叶松林、樟子松林、落叶松-白桦混交林、白桦林5种典型林分不同坡位地表细小死可燃物含水率动态进行研究,构建了不同季节防火期、不同林分地表死可燃物含水率的预测模型,并分析了其预测误差.结果表明:相同林分地表可燃物含水率在春季和秋季差异显著;在相同季节相同林分下不同坡位可燃物含水率存在差异.采用Nelson模型对地表死可燃物含水率预测的平均绝对误差(MAE)的平均值为0.13,略低于Simard模型(0.14),明显低于气象要素回归模型(0.25).Nelson和Simard模型的预测效果好于气象要素回归模型.秋季模型对地表死可燃物含水率的预测精度好于春季模型和春季秋季混合模型.
In the spring and autumn Daxinganling area Xilinji Forestry Bureau aspen - Betula platyphylla forest, Larix forest, Pinus sylvestris var. Mongolica, mixed forest of larch - Betula platyphylla, Betula platyphylla forest five kinds of typical stand at different slope surface moisture content of combustible material dynamics The prediction models of moisture content of dead combustible material in different seasons of firebreak period and forest stand were established and their prediction errors were analyzed.The results showed that the moisture content of surface combustible material in the same stand differed significantly in spring and autumn, The moisture content of combustibles at different slope positions under the same stand was different.The mean absolute error (MAE) of prediction of moisture content of dead combustibles by surface Nelson model was 0.13, slightly lower than that of Simard model (0.14) Meteorological regression model (0.25) .Nelson and Simard model prediction better than the meteorological regression model.The autumn model for prediction of surface dead combustible water content is better than the spring model and the spring autumn hybrid model.