Study On Fuel Moisture Content Prediction Of Fuels In Typical Stands, Kunming

Yunnan Province is a high fire risk region in China where accurately prediction of fuel moisture content is of great significance for the improvement of fire danger rating accuracy. Successive observation of fuel moisture contents of tree、shrub、herbaceous and dead surface fuels in9stands in Kunming, Yunnan Province were conducted in2013fire season. Dynamics and affecting factors of fuel moisture content were analyzed. Moisture prediction models were established using a vapor exchange method, FWI method and a method with mixed weather variables and FWI indexes. These models employ easily obtained weather variables from weather stations and have accuracy within the limit of similar studies, and hence suitable for fire danger rating in the region.For prediction of tree fuel moisture, no significant difference existed among the three types of models, vapor exchange models are the best considering because not all mixed models have FWI indexes come into though mixed models have the least error. The vapor exchange models have MAE7.4%-14.6%, averaged10.0%, and MRE7.0%-11.3%, averaged8.6%.For prediction of shrub fuel moisture, no significant difference existed among the three types of models*vapor exchange models are the best considering because three of the four mixed models improved a little precision though mixed models have the least error. The vapor exchange models have MAE6.6%-33.0%, averaged14.8%, and MRE5.6%-29.1%, averaged12.8%.For prediction of herbaceous fuel moisture, no significant difference existed between the vapor exchange models and mixed models, vapor exchange models are the best considering because only one mixed models have FWI indexes come into though mixed models have the least error. The vapor exchange models have MAE10.1%-37.6%, averaged19.9%, and MRE8.0%-27.8%, averaged16.6%.For prediction of fuel moisture <35%, vapor exchange models are the best considering less computation requirement though mixed models have the least error. The vapor exchange models have MAE2. l%-6.0%, averaged3.6%, and MRE11.4%-32.7%, averaged21.3%. For predicting fuel moisture content after rain, no significant difference existed among the three types of models. Considering easy computation, vapor exchange models are still the best choice with MAE8.2%-14.2%,averaged10.6%, and MRE48.7%-91.3%, averaged61.4%。FWI indexes are correlated with local fuel moisture but not as close as weather variables.