Comparing effects of fire modeling methods on simulated fire patterns and succession: a case study in the Missouri Ozarks

Abstract

We compared four fire spread simulation methods (completely random, dynamic percolation, size-based minimum travel time algorithm, and duration-based minimum travel time algorithm) and two fire occurrence simulation methods (Poisson fire frequency model and hierarchical fire frequency model) using a two-way factorial design. We examined these treatment effects on simulated forest succession dynamics and fire patterns including fire frequency, size, burned area, and shape complexity of burned patches. The comparison was carried out using a forest landscape model (LANDIS) for a surface fire regime in the Missouri Ozark Highlands. Results showed that incorporation of fuel into fire occurrence modeling significantly changed simulated dynamics of fire frequency and area burned. The duration-based minimum travel time algorithm produced the highest variability in fire size, and the dynamic percolation method produced the most irregular burned patch shapes. We also found that various fire modeling methods greatly affected temporal fire patterns in the short term, but such effects were less prominent in the long term. The simulated temporal changes in landscape-level species abundances were similar for different fire modeling methods, suggesting that a complex fire modeling method may not be necessary for examining coarse-scale vegetation dynamics.