Generation of synthetic infrared remote-sensing scenes of wildland fire

Abstract

We describe a method for generating synthetic infrared remote-sensing scenes of wildland fire. These synthetic scenes are an important step in data assimilation, which is defined as the process of incorporating new data into an executing model. In our case, this is a fire propagation model. The scenes are built using the surface output of fire position from a fire propagation code and prior knowledge of fire physics and behavior to estimate the shape of the flame. The scene radiance is then estimated by employing a physics-based ray-tracing model called DIRSIG to render the radiation that would reach a sensor on an airborne platform. Values of the Fire Radiated Energy calculated from the synthetic radiance scene compare well with previously published values, providing validation of the method.