The authors found that long term climatic stress, measured by climatic water deficit, predisposed trees to higher mortality from fire damage. The author suggest that warming temperatures increase fire severity, and ultimately tree mortality, independent of fire intensity.

Historically, high wind speeds were required to initiate crown fire behavior across all forest types within the project area. By the year 2000, common windspeeds of 45 km/h were projected to be sufficient to initiate crown fire at most sites due to changes in canopy bulk density. The increase in canopy fuels continuity has already and is expected to lead to broad-scale crown fire instead of a highly mixed pattern of burn severity that was common historically.

At the higher elevation sites, despite long fire-free periods, fire severity was consistent with historical fire regime patterns and resulted in fuel conditions closer to the historical range of variation. Furthermore, the authors found that regeneration success was not tied to fire severity, but favored fire-susceptible species, typically resprouting species. Fire effects on cool, mixed-conifer ecosystems may allow the composition and structure of these ecosystems to be more resilient to the effect of catastrophic wildfire due to climate change.

All forest types were significantly less dense in 1876 compared to 2000; however, the changes were less apparent in the higher altitude forests. The authors observed a shift in composition toward more mesic species in lower altitude forests. They suggest that because fire frequencies are typically longer in higher altitude forests, the impact of fire exclusion was not as great as low altitude forests where frequent fire thinned encroaching fire-susceptible mesic species.

The authors found that fire was highly synchronous across the region and were associated with climate conditions. Their analysis found strong relationships between dry years, preceded by one or two wetter years, with increased fire activity across the region. Specifically, widespread fire years were synchronous with negative, extreme PDSI events while fire was asynchronous when PDSI was positive. The authors also found stronger relationships with interannual climate variability than with long-term, mulidecadal climate cycles driven by the ENSO.

The study landscape was characterized by patches of ponderosa pine forest divided by a matrix of pinyon–juniper (PJ), sagebrush shrublands, and small grasslands. The authors did not find regional synchrony between the patches of ponderosa pine within the study area, suggesting that bottom-up controls, such as fuels, may have more strongly influenced the historical occurrence of fire in this region. The authors did find, however, as the scale of the analysis increased to larger regional areas, climate-driven fire synchrony increased, and fire was associated with dry condition in the year of the fire and wet conditions two years prior to the fire. The authors suggest that climate may be less influential on fire activity in ecosystems that are fuel limited.

The authors found a positive feedback between the initial severity of a fire and the severity of a later reburn, especially concerning high severity forest fire. Conversely, low- or moderate-severity fires may regulate future fires to similar severities, which follows historical patterns of frequent, low-severity fires in dry coniferous forest systems.

The authors found positive trends in the number of large fires in most ecoregions of the western U.S. including the Arizona-New Mexico Mountains with an average increase of 0.6 large fires per year. The ecoregions with trends toward increased fire activity also concurrently displayed increasing trends in drought severity over the period of the analysis. While the authors acknowledge that it is difficult to directly correlate human-caused climate change and fire activity, they assert that regardless of cause, changes in fire activity are part of larger trends tied to higher temperature and drought that are likely to continue.

The authors found that spreading fires in ponderosa pine and dry mixed-conifer forests were associated with prior wet conditions and larger fires affecting mesic forests were more strongly associated with extreme drought conditions the year of fire, historically. Currently, spreading fire only occurs during persistent drought, suggesting that fire suppression has altered fuels so that dry conifer sites are no longer fuel-limited, and fire regimes in these systems behave like more mesic mixed-conifer sites. Fire severity has also increased by four times that of previous large fires, resulting in high tree mortality and low regeneration in burn scars.

The authors’ results support the intermediate fire-productivity hypothesis across the various ecoregions of the globe. They found that increasing productivity increases fire activity until a moisture threshold is met where moisture levels then increasingly limits flammability and fire activity. Fire that occurs in moist and productive regions, where fuel is not limiting, depends on changes in climate, such as increased temperatures or periods of drought, to burn. Conversely, very arid systems are fuel limited and sensitive to changes in fuel loading and continuity.