5th Symposium on Fire and Forest Meteorology and the 2nd International Wildland Fire Ecology and Fire Management Congress

Monday, 17 November 2003
Using a regional climate model to diagnose climatological and meteorological controls of wildfire in the western United States
Steve W. Hostetler, U.S. Geological Survey, Corvallis, OR; and P. J. Bartlein, J. O. Holman, A. M. Solomon, and S. L. Shafer
Poster PDF (2.5 MB)
We are using a regional climate model, RegCM2, forced by boundary conditions (i.e., time-varying vertical profiles of temperature, wind, humidity, and sea-surface temperature) derived from global, 6-hourly NCEP/NCAR Reanalysis Project output, to examine (a) the seasonal-to-interannual variations of climate, and (b) the diurnal-to-weekly variations of weather and land-surface conditions that jointly determine the number and size of wildfires in the western United States. The model is run over a 45 km grid that covers the western U.S. and adjacent areas of Canada and Mexico. The model grid includes 3510 terrestrial grid points, and is of sufficiently high resolution to depict climatically important physiographic features (e.g. the Cascade Range, Columbia Basin and Snake-River Plain in the Pacific Northwest) that are absent in coarser-resolution GCMs. We focus here on the interval 1980 to present, for which spatially disaggregated daily data on the initiation of fires and their causes are available for the region.

RegCM2 provides a number of simulated and derived atmospheric variables and indices that describe and relate the role of atmospheric circulation, moisture flux, and instability in triggering wildfires. The model also includes a sophisticated land-surface physics package, LSX, that calculates explicitly surface water- and energy-balance variables that can be related to the propensity for fires to occur. We are also using several off-line terrestrial ecosystem models to investigate the influence of both daily and seasonal variations of weather and climate on the production and decomposition of biomass, and hence fuels.

Examination of monthly averages and anomalies of the RegCM2 output reveals the specific atmospheric and surface water- and energy-balance anomalies that determine the characteristics of individual summer fire seasons. Although there are significant spatial and temporal variations in the particular anomalies that favor severe fire seasons, the generally high-fire years 1988, 1996, 2000 and 2002, for example, have in common widespread and strongly negative cloudiness, soil-moisture, and surface relative humidity anomalies, and positive sensible heating anomalies, all of which are the direct result of large-scale atmospheric circulation anomalies that resulted in persistent subsidence over the region.

On the daily time scale, the RegCM2 simulations reveal the influence of atmospheric stability, mid-tropospheric moisture flux and upper-level circulation on widespread outbreaks of fire. For example, during the interval 12 -16 August 1996, nearly 2000 lightning started fires occurred across the western U.S. (with over 1000 starting on 13 and 14 August). Steep 700-500 hPa environmental lapse rates and highly negative lifted-index values prevailed over this interval as a consequence of strong surface heating and advection of subtropical moisture. This unstable synoptic situation was modulated by a weak upper-level trough that moved onshore from the eastern Pacific and supported numerous “dry” convective events.

Supplementary URL: http://geography.uoregon.edu/fireclim/