Matthew G. Rollins

Mapping fire regimes across time and space: Understanding coarse and fine-scale fire patterns

This paper was presented at the conference ‘Integrating spatial technologies and ecological principles for a new age in fire management’, Boise, Idaho, USA, June 1999 Maps of fire frequency, severity, size, and pattern are useful for strategically planning fire and natural resource management, assessing risk and ecological conditions, illustrating change in disturbance regimes through time, identifying knowledge gaps, and learning how climate, topography, vegetation, and land use influence fire regimes. We review and compare alternative data sources and approaches for mapping fire regimes at national, regional, and local spatial scales. Fire regimes, defined here as the nature of fires occurring over an extended period of time, are closely related to local site productivity and topography, but climate variability entrains fire regimes at regional to national scales. In response to fire exclusion policies, land use, and invasion of exotic plants over the last century, fire regimes have changed greatly, especially in dry forests, woodlands, and grasslands. Comparing among and within geographic regions, and across time, is a powerful way to understand the factors determining and constraining fire patterns. Assembling spatial databases of fire information using consistent protocols and standards will aid comparison between studies, and speed and strengthen analyses. Combining multiple types of data will increase the power and reliability of interpretations. Testing hypotheses about relationships between fire, climate, vegetation, land use, and topography will help to identify what determines fire regimes at multiple scales.

MAPPING FUELS AND FIRE REGIMES USING REMOTE SENSING, ECOSYSTEM SIMULATION, AND GRADIENT MODELING

Maps of fuels and fire regimes are essential for understanding ecological relationships between wildland fire and landscape structure, composition, and function, and for managing wildland fire hazard and risk with an ecosystem perspective. While critical for successful wildland fire management, there are no standard methods for creating these maps, and spatial data representing these important characteristics of wildland fire are lacking in many areas. We present an integrated approach for mapping fuels and fire regimes using extensive field sampling, remote sensing, ecosystem simulation, and biophysical gradient modeling to create predictive landscape maps of fuels and fire regimes. A main objective was to develop a standardized, repeatable system for creating these maps using spatial data describing important landscape gradients along with straightforward statistical methods. We developed a hierarchical approach to stratifying field sampling to ensure that samples represented variability in a wide variety of ecosystem processes. We used existing and derived spatial layers to develop a modeling database within a Geographic Information System that included 38 mapped variables describing gradients of physiography, spectral characteristics, weather, and biogeochemical cycles for a 5830-km 2 study area in north- western Montana. Using general linear models, discriminant analysis, classification and regression trees, and logistic regression, we created maps of fuel load, fuel model, fire interval, and fire severity based on spatial predictive variables and response variables measured in the field. Independently evaluated accuracies ranged from 51 to 80%. Direct gradient modeling improved map accuracy significantly compared to maps based solely on indirect gradients. By focusing efforts on direct as opposed to indirect gradient modeling, our approach is easily adaptable to mapping potential future conditions under a range of possible management actions or climate scenarios. Our methods are an example of a standard yet flexible approach for mapping fuels and fire regimes over broad areas and at multiple scales. The resulting maps provide fine-grained, broad-scale information to spatially assess both ecosystem integrity and the hazards and risks of wildland fire when making decisions about how best to restore forests of the western United States to within historical ranges and variability.

Landscape-scale controls over 20th century fire occurrence in two large Rocky Mountain (USA) wilderness areas

Topography, vegetation, and climate act together to determine thespatial patterns of fires at landscape scales. Knowledge oflandscape-fire-climate relations at these broad scales (1,000s hato 100,000s ha) is limited and is largely based on inferences andextrapolations from fire histories reconstructed from finer scales. In thisstudy, we used long time series of fire perimeter data (fire atlases) and datafor topography, vegetation, and climate to evaluate relationships between large20thcentury fires and landscape characteristics in two contrastingareas: the 486,673-ha Gila/Aldo Leopold Wilderness Complex (GALWC)in New Mexico, USA, and the 785,090-ha Selway-BitterrootWilderness Complex (SBWC) in Idaho and Montana, USA. There were importantsimilarities and differences in gradients of topography, vegetation, andclimatefor areas with different fire frequencies, both within and between study areas.These unique and general relationships, when compared between study areas,highlight important characteristics of fire regimes in the Northern andSouthernRocky Mountains of the Western United States.Results suggest that amount and horizontal continuity of herbaceous fuels limitthe frequency and spread of surface fires in the GALWC, while the moisturestatus of large fuels and crown fuels limits the frequency of moderate-to-highseverity fires in the SBWC. These empirically described spatial and temporalrelationships between fire, landscape attributes, and climate increaseunderstanding of interactions among broad-scale ecosystem processes. Resultsalso provide a historical baseline for fire management planning over broadspatial and temporal scales in each wilderness complex.

Evaluating a century of fire patterns in two Rocky Mountain wilderness areas using digital fire atlases

Changes in fire size, shape, and frequency under different fire-management strategies were evaluated using time series of fire perimeter data (fire atlases) and mapped potential vegetation types (PVTs) in the Gila – Aldo Leopold Wilderness Complex (GALWC) in New Mexico and the Selway–Bitterroot Wilderness Complex (SBWC) in Idaho and Montana. Relative to pre-Euro-American estimates, fire rotations in the GALWC were short during the re cent wildfire-use period (1975–1993) and long during the pre-modern suppression period (1909–1946). In contrast, fire rotations in the SBWC were short during the pre-modern suppression period (1880–1934) and long during the modern suppression period (1935–1975). In general, fire-rotation periods were shorter in mid-elevation, shade-intolerant PVTs. Fire intervals in the GALWC and SBWC are currently longer than fire intervals prior to Euro-American settlement. Proactive fire and fuels management are needed to restore fire regimes in each wilderness complex to within natural ranges of variability and to reduce the risk of catastrophic wildfire in upper elevations of the GALWC and nearly the entire SBWC. Analyses of fire atlases provide baseline information for evaluating landscape patterns across broad land -

Evaluation of novel thermally enhanced spectral indices for mapping fire perimeters and comparisons with fire atlas data

We evaluated the potential of two novel thermally enhanced Landsat Thematic Mapper (TM)‐derived spectral indices for discriminating burned areas and for producing fire perimeter data (as a potential surrogate to digital fire atlas data) within two wildland fires (1985 and 1993) in ponderosa pine (Pinus ponderosa) forests of the Gila Wilderness, New Mexico, USA. Image‐derived perimeters (manually produced and classified from an index image) were compared to fire perimeters recorded within a digitized fire atlas. For each fire, the highest spectral separability was achieved using the newly proposed Normalized Burn Ratio‐Thermal (NBRT1) index (M = 1.18, 1.76, for the two fires respectively). Correspondence between fire atlas and manually digitized fire perimeters was high. Landsat imagery may be a useful supplement to existing historical fire perimeters mapping methods, but the timing of the post‐fire image will strongly influence the separability of burned and unburned areas.