|
Accurately predict fuel moisture and is the key to do a good job against fire weather and fire behavior prediction method based on the time delay and the equilibrium moisture content of the most widely used applications, including Catchpole, et al (2001) is a result of Nelson (1984) semi- physical model and universal good future. But there are some problems in the method there is no in-depth study: ① The Catchpoleet al (2001) modeled the same data and validation data, naturally increase the accuracy of the model. If different from modeling data validation data, how the accuracy of the model is not yet clear. Of ② modeling data length of the parameter estimates and projections results is unclear. The equilibrium moisture content of the response function of the temperature and humidity (3) the method using Nelson (1984) model. There are four models (Viney, 1991; Liu Xi, 2007), which, Simard (1968) model is used by the U.S. National Fire Danger Rating System (NFDRS), (Xi Liu et al, 2007) show that the model intends Nelson (1984) model effects of the combined effect than Therefore, if using these equilibrium moisture response model, the effect of the method are also unclear. In this paper, the study of continuous observation of the different sizes of fuel moisture and environmental factors on these three issues. On this basis, the application Catchpole, et al (2001) method of birch (Betulaplatyphylla), hardwood (Hard-width), Quercus mongolica (Mongolian oak), 2 (Fraxinusmandshurica), shrub (Shrubs) and grass hardwood Austin (Meadow), six forest litter, half humus, humus and mixed combustible Delay and balance the moisture content was estimated, and the moisture in the forecast. The results showed that: 1) with modeling data is different from the verification data, but the length of the modeling data is large (at least 84), Catchpole, et al (2001) method still has a high degree of accuracy, that the method in the use of field observation data to predict the moisture content has strong applicability. But this time modeling larger amount of data, it is recommended that more than 80. Modeling data about the length of 30 the error standard less demanding (3%), can also be applied. 2) when the When modeling data length is less, Nelson (1984) Model-based moisture prediction error is to be significantly less than that Simard (1968) model, when the modeling data is large (more than 84), two models predictive effect difference little. 3) Stand on the time delay and the equilibrium moisture content has little effect; decomposed Delay significant, at a temperature of 20 ° C, significantly affect the equilibrium moisture content in the range of 25% to 45% relative humidity. 4) different forest types moisture forecast error difference is obvious; different the decomposed moisture prediction: Litter Half humus |