This study compared fire behavior and emissions predictions made from out-of-the-box LANDFIRE data to those made with field-based on-the-ground data. For forested vegetation, all four study sites showed significantly greater fire behavior and emissions when using LANDFIRE data, which can mostly be attributed to differences in forest and fuel characteristics mapping. Local data is generally considered more appropriate for projects of the scale examined here, but acquiring field based data is costly and time consuming. Additionally, application of plot-based data to landscapes is an important consideration in assessing the accuracy of these types of simulations. Modification of LANDFIRE inputs based on field data may be the most desirable alternative at this point. As the accuracy of remotely sensed inputs improves, their use at smaller scales may become more viable.
