Twenty-two fires in a laboratory wind tunnel and 8 field fires were studied with video techniques to determine relationships between their flame characteristics and fire behavior. The laboratory fires were in pine needle fuel beds with and without an overlying stratum of live vegetation. These fuels simulated 2-year roughs in southeastern fuel types. The field bums were in 1- and 2-year roughs in similar fuels. Byram's fire intensity ranged from 98 to 590 kW/m in the laboratory, and from 355 to 2755 kW/m in the field. Flame lengths were proportional to the square root of fire intensity when fuel consumption exceeded 0.5 kg/m^2, in agreement with predictions from buoyant flame theory. However, for burns in the needle layer (consumption approximately 0.5 kg/m^2), flame lengths were constant at about 0.5 m, regardless of intensity. Similar values were observed on two of the field fires. It is speculated that flame length is limited by a boundary layer pattern for the overall flow, even though the flames themselves did not exhibit boundary layer characteristics. Also, laboratory correlations of flame tilt angle and fire intensity with other fire and weather variables depart from buoyant flame theory. Further study under field conditions is needed before relationships involving flame tilt angle, fire intensity, and wind speed should be used in practical applications.

[This publication is referenced in the "Synthesis of knowledge of extreme fire behavior: volume I for fire managers" (Werth et al 2011).]