Fire Risk

Placing weather-related events in a historical context can be a useful exercise, especially when trying to understand the meteorological conditions that certain hazards, e.g. wildfires and drought, favor and how these events may be exacerbated by changing climate.

Wildfires have received considerable attention over the past two years, due to their devastation (Camp and Carr Fires in California; Substation Fire near The Dalles) or economically or socially valued location (Eagle Creek Fire in the Columbia River Gorge and Chetco Bar Fire in southwestern Oregon).

Statistical analysis shows that warm, dry summers are associated with higher area burned (McKenzie et al 2004, Westerling et al., 2006). Large fires increased in the western US from 1984-2011 in a warming climate (Dennison et al., 2014) and human-caused climate change was responsible for the increase in area burned in forests in the western US from 1984-2015 (Abatzoglou and Williams, 2016).

Fire Season

Fire season in Oregon runs roughly from late July to mid-September, though it can start earlier and end later, as was the case in 2018. Fire activity is dependent on many anthropogenic and natural variables, and warmer or drier seasons can create conditions favorable for wildfires. In Figure 8 we define fire season using a rough definition of the months of July-September. The upper left quadrant represents the warmest and driest years in this historical record; the lower right quadrant shows the wettest and coolest years. Years for the climate analysis are only labeled from 2002-2018, the same period of record as the fire data, to reduce clutter. Circles are meant to show the acres burned in each of these years, binned into groups.

The fire seasons with the most acres burned (2012, 2014, and 2017) are among the warmest in the record, and notably warmer than most of the other years in the upper left quadrant. July-September 2012 and 2017 were drier than normal, but the same time period in 2014 was slightly wetter than the historical average. There are years (2003, 2009) that were in the top 10 warmest July-Septembers on record, but had relatively small areas of wildfire activity. 2002 and 2012 had significant large fires (over 500,000 acres) boosting the overall total; the 2002 Biscuit Fire in southwestern Oregon and the 2012 Long Draw fire in eastern Oregon.

Both fire seasons topped 1,000,000 wildland acres burned. 2010 had near-normal precipitation and temperature and 2004 was wetter than normal for the fire season; these two years had the smallest area burned on record. In this typically arid season in Oregon, a few rainfall events can skew the entire season.

Impact

2013 was notable as it was Oregon’s wettest September on record, owing to two large storms early and late in the month. July-September has consistently been warmer and drier than the 1895-2018 average for most of the past 17 years, creating prime conditions for potentially large wildfire seasons. And while large wildfire seasons can be dominated by single events, warmer conditions tend to be more favorable for an above average area burned in Oregon, consistent with the rest of the western US. None of the years with above-average area burned were near normal in seasonal temperature and precipitation. In a changing climate, fire activity in Oregon will continue to be influenced by warming temperatures and longer fire seasons. Projections using vegetation models (previously published in OCAR3, Dalton et al. 2017) highlighted spatial differences in changing fire risk, emphasizing that more frequent fires could be expected even in the wet western third of the state, and indeed as the Eagle Creek fire (summer 2017) showed, that prediction is coming true.

everett-mcintire-789079-unsplash (1).jpg

Weather data can be used to calculate fire risk in various ways. Operational agencies often use the energy release component (ERC) as well as measures of fuel moisture and wind speeds. One measure of the fuel moisture is the ‘100-hr’ fuels moisture, which is the amount moisture within vegetation (the ‘fuel’), averaged over 100 hours. Figure 9 shows the ‘100-hour’ fuel moisture, specifically the number of days per summer when the fuel moisture is below the 3rd percentile (“extreme”). The largest increases in the frequency of extreme fire risk are in the eastern third of Oregon and in the Willamette Valley.