It is common to observe that seedlings and other vegetation often 

 do not become established for several years where large slash piles 

 have been burned. A profusion of vegetation may fringe the area, but 

 the center may be bare. 



These efforts also should provide inputs to hydrologic and soil stability impacts. 



SIMULATING FIRE BEHAVIOR 



We are concerned only with nonuniform fire spread in the horizontal plane. 

 Arrays with large vertical nonuniformity are not considered. The results of the 

 simulation must reflect the variability in fire behavior due to the spatial non- 

 uniformity of the fuel. Fire spread results are presented as a frequency distri- 

 bution of the rate of spread of the fire front at variable lapsed times from igni- 

 tion of the initial fire front. Intensity distributions are presented for those 

 lapsed times . ^ 



The simulated fuel bed consists of an array of cells. Each cell is described 

 by a set of basic fuel parameters (Rothermel 1972) that affect the rate of fire 

 spread. The time to move from cell to cell depends on the parameters of the two 

 adjacent cells. The fuel is assumed to be evenly distributed within each cell. 



The fire begins as a line source and travels from cell to cell by contagious 

 growth through a series of ignitions and spreads at a rate based on a minimum delay 

 time Cappendix 1) within each individual cell (fig. 1). Hexagonal cells are used 

 because they do not have point contacts when arranged in an array and offer the 

 maximum number (six) of growth directions while maintaining a constant distance 

 between cell centers. 



The delay time is the core of this analysis. Delay time is viewed as the time 

 it takes for the fire to spread through the cell. Thus it is the amount of time the 

 fire is delayed in a cell before it can attack an adjacent cell. The delay time is 

 the sum of two parts: the time to achieve the quasi-steady spread rate in the cell 

 and the time to spread the remaining distance at a steady rate of spread. The time 

 to reach the quasi-steady state is assumed to be the residence time of the cell and 

 the remaining time is the remaining distance divided by the quasi-steady rate of 

 spread for that cell. Consequently, the cell chosen must be of a size such that the 

 time for the fire to spread through the cell is equal to or greater than the resi- 

 dence time of the fire in the cell. An essential restriction of the model is that 

 the influence from a burning cell extend no further than its immediate neighbors. 



After passage of the front, the fire will remain burning in the cell for a 

 period of time dependent on cell fuel properties, the properties of the cell it 

 was ignited by, and the size of the fuel cell. The delay time can be shortened 

 after ignition if heat from another adjacent cell has sufficient intensity to 

 offer a delay time less than the present waiting time. 



Fire from one cell can ignite an adjacent cell after it has reached its es- 

 cape level, i.e. waited a time equivalent to its delay time. If the adjacent cell 

 is presently unignited, the fire moves in, delay time is assigned, and the fire be- 

 gins its waiting period to reach its escape level. If the adjacent cell is already 

 ignited and the proposed delay time is less than the present waiting time, then the 

 waiting time is replaced by the delay time. Otherwise the cell retains its waiting 

 time. A detailed discussion of the delay time appears in appendix 1. 



4 



