Predicting spatio-temporal variability in fire return intervals using a topographic roughness index

The shapes of landscapes are fundamental to ecosystem processes at various spatial scales. Topographic roughness index (TRI) is a measure of variability in the landscape surface and a proxy of the potential of disturbances to propagate across the earth's surface, such as a wildland fire burning across a landscape. We describe the significance of TRI, present methods for calculation, and demonstrate the utility of the index in a fire frequency prediction model. The model was used to show how the relationships between topography, fire, and humans changed during the period of AD 1620-1850 for a study area (5180 km 2) in Missouri, USA. The model predicted historic mean fire return intervals from TRI and two human population variables. The model explained 46% of the variation in mean fire return intervals and demonstrated that topographic roughness was most important in controlling fire frequency during the period AD 1620-1780 when human population density was lowest (<0.35 humans/km(2)). Due to increases in human population, mean fire return intervals were shortened by LIP to one-fourth of their original length and the landscape became more homogeneous with respect to fire frequency despite topographic roughness. The use of TRIs in wildland fire research aid in quantifying and visualizing topographic variability and could be applicable to multiple scales and ecosystem processes. (c) 2007 Elsevier B.V. All rights reserved.