This thesis evaluates the effectiveness of BEHAVE: Fire Behavior Prediction and Fuel Modeling System in predicting fire behavior in the Northeastern pine-oak forest. This fuel complex is composed primarily of a litter and huckleberry shrub understory with a pitch pine and oak overstory. Measurements of fuel bed physical characteristics, weather and fire behavior are taken from a series of prescribed burn studies in Cape Cod National Seashore in Massachusetts. Site-specific fuel models are constructed which provide the necessary inputs for fire predictions. Observed spread rates and flame lengths are over-predicted by BEHAVE for burns conducted during the winter (dormant season) and under-predicted for burns conducted during the summer (growing season). Attempts to improve winter predictions are successful when the litter moisture is adjusted in order to account for the live wintergreen which increases the overall moisture content of the surface fuels. A sensitivity study is performed where each input parameter is varied over a reasonable interval in order to view its impact on predictions. The model’s high sensitivity to fuel bed depth and 1-hr surface-area-to-volume ratio appear to be the cause for fire prediction deviations during the winter, while the high live fuel moisture contents appear to overwhelm and suppress fire behavior predictions during the summer. It is concluded to overwhelm and suppress fire behavior predictions during the summer. It is concluded that BEHAVE’s representation of fuel complexes as a homogeneous fuel bed with constant properties does not take into account the unique features of the litter and shrub components. An alternative, simple model of fire spread is developed which treats each component as a separate fuel bed. The model is based on a measurement of the heat release rate which can be determined directly through the principle of oxygen consumption calorimetry. Future work using small- and large- scale testing apparatus will help determine the ignition process of the live shrubs and the effect of parameters such as moisture content on the burning characteristics of the fuels.

Creator

• Dell'Orfano, Michael E

Contributors

• Dembsey, Nicholas A.

Degree

• MS

Unit

• Fire Protection Engineering

Publisher

• Worcester Polytechnic Institute

Language

• English

Identifier

• etd-10264-080257

Keyword

• Fire
• forest
• fire behavior
• Modeling

Advisor

• Dembsey, Nicholas A.

Defense date

• 1996-08-01

Year

• 2004

Date created

• 2004-10-26

Resource type

• Thesis

Rights statement

• In Copyright