Climate is a major driver of regionally synchronous fires in many regions of the US. Regional fire events, like those that occurred in the Northern Rockies in 1910, 1988, 1994, 2000 and 2002, typically occurred during years when drought was similarly extensive, and account for the majority of area burned. During such years, the threats to people and their property are highest, because fires during these years can quickly overwhelm our ability to suppress them. Furthermore, regional fire events play a critical role in governing ecosystem dynamics at broad scales. Predicting the climate conditions under which these ecologically and socially important regional fire years occur would have major benefits for fire management in the US.

The Joint Fire Science Program has funded this 3-year research project to identify the climate drivers of regional fire and fuel dynamics in the Northern Rockies in the past, present, and future. We will identify regional fire years from two sources: multicentury tree-ring reconstructions and multidecadal fire atlases (Table 1 below).

To elucidate the climate forcing of past fires, we will reconstruct the occurrence of regional fire years from synchrony in fire-scar dates among at least 15 widely separated sites in the region. While multicentury histories of fire have been reconstructed from tree rings for a number of sites in the Northern Rockies, none of these studies used crossdating to obtain the annually accurate records of fire occurrence that are essential for identifying individual years during which fires were regionally synchronous, and hence for elucidating the forcing of annually varying climate processes. To elucidate present, i.e., twentieth-century, climate forcing of fires, we will compile previously digitized fire atlases from at least 11 National Forests in the region. In spite of their potential for identifying synchronous and extensive regional fire years, fire atlases have not yet been used to assess regional fire-climate relationships. To elucidate effects of future climate forcing of fire, we will use simulation modeling. We will parameterize a landscape simulation model using information derived from the fire scars and fire atlases about past and present fire regimes. We will also simulate different fuels management scenarios, focusing on the implications for large fire years and the degree to which fuels management will be effective under those conditions.

Increasingly, it is possible to predict climate for future fire seasons, so our information will help fire managers anticipate when fire management must focus primarily on fire suppression versus fuels management, and where such fuels management is likely to affect fire behavior during future regional fire years.

This three-year study, funded by the Joint Fire Sciences Program (2003-2006), has four main objectives. These are to:

1. Elucidate the climate forcing of past and present regional fire years in the Northern Rockies.
2. Assess the agreement between fire atlas and fire scar data at multiple scales.
3. Determine the probability of future regional fire years in the Northern Rockies under different climate scenarios.
4. Determine the effect of present fuel and fire management on future regional fire years.

Beginning in the summer of 2004, we will sample approximately 15-20 fire scar sites in western Montana, Idaho, and extreme eastern Oregon. These sites will be distributed evenly over the northern Rocky Mountains. At each site, we will sample ten to twenty fire scarred trees, logs or stumps. Using dendrochronological techniques (i.e., crossdating), individual fire scars will be dated and used to reconstruct the occurrence of regional fire years from synchrony in fire-scar dates. We will compare these regional fire years to annually and decadally varying climate processes that affect weather in the northern Rocky Mountains, such as El Niño/Southern Oscillation and the Pacific Decadal Oscillation. Additionally, sites for a gridded analysis will be selected and sampled to generate fire history and stand age reconstructions in areas covered by the fire atlases. Modern fire atlases will be compared with modern climate data to elucidate present climate forcing of fire events.

To create a single 20th century fire atlas to represent the study area, we will compile digital fire atlas data from a minimum of 11 National Forests in the northern Rocky Mountains. The northern Rocky Mountain fire atlas will allow us to identify regional fire years, which will be defined in three ways: total area burned will be used to identify the largest fire years; geographically widespread fires will be determined based on latitude and longitude ranges of recorded fire events in a given year; and biophysically widespread fires will be identified through spatial analysis of the range of sites experiencing fire events in a given year. By identifying years with a large area burned, geographically widespread, and biophysically diverse regional fires and assessing annual climate data for these years, we can examine fire-climate relationships and synchrony for the entire northern Rocky Mountain region.

Simulation modeling will be used to investigate the consequences of climatically controlled regional fire events on future landscape patterns of vegetation and fuels. We will use the fire frequency and fire size that we derive from the fire scar and fire atlas data to parameterize a landscape simulation model. We will then use the model to conduct multiple simulation experiments on at least four landscapes that are representative of the region. In these experiments, the frequency of large fire years will be varied in combination with different assumptions regarding fuel abundance and treatments. We will summarize our results in terms of the frequency of large fires and the efficacy of fuels treatment within different vegetation types and under different climatic conditions.