| Because of relatively dry climate conditions,especially in the seasons of winter and spring,forest fuel is so dry that fires ferequently occurred in Beijing.The major forest types of Beijing is coniferous plantation,most of the understor:ies are weeds and shrub,which have little ability to withstand the forest fire,at the same time forest has the great higher forest fire danger with the frequent human activities such as tourism.In order to get the ideal prediction results about the moisture content of land surface dead combustible fuel under different site conditions,we need to establish the corresponding prediction model.Therefore,based on different types of fuel moisture,accurately choose to environmental factors,and then establish the relevant model for predicting water content in fuel,has a very important practical significance.This paper take the fuel moisture content of five tree species in Yunxiu ting and Luo ba region of Miaofeng mountain as study objects,a total of ten forest types in three kinds of time delay(1 time delay,10 time delay,100 time delay).We take daily observation in Beijing spring fire period from 2016.4.2 to 2016.5.31,and analysis the factors which affect the different kinds fuel moisture content in six forest types,including Quercus variabilis in shady slope,Quercus variabilis in sunny slope,Acer mono Maxim in shady slope,Pinus tabuliformis Carriere in shady slope,Platyeladus orientalis in sunny slope and shady slope,establish the index of stepwise regression and nonlinear regression model for predicting fuel water content seperately,the main results are as follows:(1)Analyzed the dynamic change of moisture content and environmental factors of different tree species and different kinds of the surface dead fuel under the condition of the different slope direction(shady slope and sunny slope)and different slope positions(uphill and downhill)in spring fire prevention period.The rule of fuel moisture content in specific type changed with different slope positions is disaffinity.In general,the variations of moisture content of the branch in 1 time delay is much higher than the content of deadwood 10 delay and 100 delay.(2)The correlation analysis of the fuel moisture content and environmental factors(environmental factor at present and environment factor in earlier stage)show that fuel moisture content is negatively correlated with temperature,5oil temperature and wind speed;the moisture content were positively correlated w:ith the relative humidity.The impact of different environmental factors on different kinds of fuel moisture content have different significance.(3)The fuel moisture content prediction model is based on stepwise regression analysis,when the significant level is 0.05,R2>0.5,all pass on inspection.The MAE of moisture content between measured and predicted value is 0.19%?1.49%,MRE is 2.01%?13.21%,<15%,the effect of model is great.(4)The exponential equation is based on nonlinear regression equation,the 18 prediction model of R2 were greater than 0.5,the MAE of moisture content between measured and predicted values is 0.19%?1.47%,MRE is 1.62%?12.48%,<15%,the effect is so good that the model can put good into application.(5)Through the comparing with the fit of the two kinds of prediction model,a conclusion is draw that:stepwise regression ruodel have a great effect on the deadwood of Quercus variabilis in shady slope with 1 time lag,deadwood of Quercus variabilis in sunny slope with three type of time lag,deadwood of Pinus tabuliformis Carriere in shady slope with 100 time lag,deadwood of Platycladus orientalis in sunny slope with land 10 time lag and shady slope with three type of time lag.However,other types of fuel moisture content prediction is more suitable for nonlinear exponential model. |
PDF Full Text Request