Prediction Of Moisture Contents Of Typical Fuels In Nanchang, Jinagxi Province:Optiomation Of Models And Evaluation Of FWI

Subtropical forest is an important vegetation type in south China, where fires occur frequently. Better prediction of fuel moisture content in the region is of benefit to improve fire danger rating. Dead and live fuels under5typical stands in Chayuanshan Forest Farm, Nanchang, Jiaxi Province were selected and the dynamics of moisture and weather variables were observed. Factors affecting fuel moisture content were analyzed. Prediction models were established using solely weather variables, solely FWI indexes and the two together as predictive variables, respectively based on dataset of different moisture range. Results show that dead fuel moisture in the region in the study period ranged from11.8%to276.6%. One fourth of them were<35%which is ready to burn and the rest have lower combustibility, which suggests low mean fire hazard in the region. Even so, since the minimum moisture is below10%, it still has the potential to burn largely. Results show that live fuel moisture in the region in the study period ranged from13.19%to278.19%. Two fifths them were<130%which is ready to burn and the rest have lower combustibility, which suggests low mean fire hazard in the region. Even so, since the minimum moisture is below13%, it still has the potential to burn largely.For the three types of dead fuel moisture prediction models established, the FWI model had the largest error, higher than the vapor exchange models and the mixed model, where there was no significant difference between the rest two models. Considering simple computation requirementt, vapor exchange models are the best choice for predicting fuel moisture in the region, for all different moisture ranges. For the dead fuel moisture, the MAE is2.25%-4.67%with a mean of3.73%, MRE11.14to23.73%averaged18.63%. the three types of live fuel moisture prediction models established, the FWI model had the largest error, some prediction models were not established using solely FWI indexes,higher than the vapor exchange models and the mixed model, where there was no significant difference between the rest two models. Considering simple computation requirementt, vapor exchange models are the best choice for predicting fuel moisture in the region, for all different moisture ranges, the MAE is3.35%-6.35%with a mean of5.13%, MRE2.63%-5.60%%averaged4.66%.The FWI indexes were closely correlated with fuel moisture in the region which indicates that they were suitable for the region and for creating a nationally unique prediction model. But systematically modification should be conducted rather than the simple one done here for higher accuracy.