Authors: Alex Jonko; Rodman Linn; Isabelle Runde; Judith Winterkamp; Russell Parsons; Carolyn SiegIn most fire spread models, slope is a factor determining fire behavior. Empirical models typically allow users to supply a local slope, which impacts fire spread rates. Nonlocal changes in slope, including topographic curvature, are inherently ignored. Using physical, self-determining coupled fire-atmosphere models, such as FIRETEC, topographic influences on fire spread can be simulated to enhance our understanding of fire behavior and processes governing it. Modeling topography can increase the computational cost of simulations, since much taller domains are required to model winds over the topography. When only local slope matters, fires can be simulated in coupled fire-atmosphere models by aligning the computational domain with the slope. This approach simulates a long, isolated slope with ambient winds parallel to the hillside, and affords great computational savings. However, the assumption of wind alignment with slope, as used in operational models, could cause unrealistic fire behavior under some scenarios. We compare a set of FIRETEC simulations with winds parallel to the surface and no curvature in the topography to simulations with upwind and downwind curvature, where winds are not aligned with slope upwind of the fire. We aim to identify potential ramifications of ignoring the effects of curvature and evaluate the generality of determining fire spread rates based on a 2D slope.