大兴安岭地表细小死可燃物含水率预测模型

Prediction Models of Fine Dead Fuels Moisture Content for Typical Stand in Daxing'an Mountains

以大兴安岭地区南瓮河保护区落叶松林(Larix gmelinii)、蒙古栎林(Quercus mongolica Fischer)、落叶松-白桦混交林(Mixture of Larix gmelinii and Betula platyphylla)(阴坡、阳坡)、沟塘草甸等4种典型林分为研究对象,运用气象要素回归法,对春季防火期和秋季防火期内的地表细小死可燃物含水率动态进行测定,构建了不同防火期、不同林型地表死可燃物含水率的预测模型,分析了相应模型的预测误差。结果表明:同林型地表可燃物含水率在春季防火期和秋季防火期差异显著;在秋季防火期,5个典型林型的地表死可燃物含水率预测平均绝对误差为0.167,平均相对误差为0.218,低于春季防火期模型和春季-秋季混合模型;秋季防火期模型对可燃物含水率预测效果最好。气象要素回归法适用于南瓮河保护区典型林型地表死可燃物含水率预测。

A study was conducted on the dynamics of the moisture content of ground surface fine dead fuels under five stand types in spring and autumn prohibitive period of forest fire,and the prediction models were established.The fuels moisture content of with Pinus sylvestnis,Quercus mongolica Fische,mixture of Larix gmelinii and Betula platyphylla（South and Nouth-facing slope）,and meadow in wetland in Nanwenghe National Nature Reserve in the Daxing＇an Mountains,the moisture content prediction models of different stand types in different prohibitive period of forest fire were established by using the meteorological element regression method.The fuel moisture content in the same stand varied with prohibitive period of forest fire.The average mean absolute error（MAE） and root mean square error（RMAE） of five stand types in autumn model were 0.167 and 0.218,respectively,lower than those in spring model and spring-autumn mix model.The accuracy of the model in autumn is generally higher than that of spring model and spring-autumn mix model,and the forecasting performance was the best.The prediction methods of the fuels moisture content based on one-day time step were applicable for the typical stand in the Daxing＇an Mountains.