COS 184-10 - Pyromes of the US: Delineating fire regimes

Friday, August 11, 2017: 11:10 AM
B116, Oregon Convention Center
Megan E. Cattau1,2, Adam Mahood1,2, Kai Kresek1, Nathan Mietkiewicz1,2, Carol A. Wessman1,3,4 and Jennifer K. Balch1,2, (1)Earth Lab, University of Colorado-Boulder, Boulder, CO, (2)Geography, University of Colorado-Boulder, Boulder, CO, (3)CIRES, University of Colorado-Boulder, Boulder, CO, (4)Ecology & Evolutionary Biology, University of Colorado-Boulder, Boulder, CO

Fire is often defined by fire regimes, described by the component characteristics (e.g., fire return interval, size, severity), which vary across space and time. Fire, controlled by either moisture or fuel, is expected to increase in frequency and severity in coming decades due to climate change (affecting moisture) and land cover / land use (LCLU) change (affecting fuel). Furthermore, climate and LCLU change projections are not uniform across the US, thus creating a patchwork of potential interactions between fire regime characteristics and the factors that affect them. Our objectives are to 1) delineate ‘pyromes,’ or areas that share relatively homogenous fire regime characteristics across the entire US and 2) evaluate if these pyromes correspond to vegetation and climatic zones. We compile a suite of fire regime characteristics, including fire intensity, severity, size / extent, return interval, ignition type, and seasonality derived from a variety of remotely sensed data. We also use novel algorithms to derive spread rate and duration. A hierarchical clustering approach is used on these fire characteristics to delineate pyromes. Pyromes are projected into geographic space and compared with vegetation and temperature- and moisture-based climate zones.


This project uses the variety of fire regime characteristics that can be derived from remotely sensed data to delineate pyromes across the US, thus taking advantage of the rapid advances in technology that greatly increase our ability to answer questions that require large spatiotemporal datasets and complex geospatial analysis. Using a hierarchical clustering approach to identify pyromes, we present a classification scheme with a variable number of pyromes depending on the hierarchical level, ranging from general (fewer pyromes with a relatively large spatial extent) to more detailed (more pyromes with a relatively small spatial extent). Regardless of the hierarchical level, the relative importance of each fire regime characteristics varies from one pyrome to another. The extent to which these pyromes map onto existing vegetation climate zones depends upon the hierarchical level. These pyromes function as a spatial framework for research on and management of fire processes, including targeting efforts to prevent and control fires and predicting how fire regimes will change in a given area under climate and land use changes.