Visualize: Insights that power innovation

California burning: An analysis of the summer 2015 wildfires

By Tomas Girnius and Arindam Samanta September 7, 2016

With at least seven lives lost and thousands of properties destroyed from raging wildfires, the summer of 2015 shaped up to be a season of terrible devastation for Northern California. Large loss-causing fires, which included the Valley, Butte, and Rocky wildfires, burned more than 216,000 acres and destroyed or damaged more than 3,000 structures.

Both Fires Occurred in Areas of High Loss Potential

FireLine Valley and Butte Fires, California map areas of high loss cost
Source: AIR Worldwide

Using industry-recognized wildfire risk management tools and models, we examined the Valley and Butte fires through the lens of a historical analysis of all fires occurring in California since 1990. Our objective was to answer several key questions: How bad was the summer of 2015 in terms of wildfires? What risk factors made the fires so challenging? And are we seeing a worsening trend?

Summer 2015 wildfire severity

Three factors influence the impact of wildfire risk on property losses: fuel, slope, and access. The primary risk factor is a property’s proximity to vegetation—grass, trees, and scrub or brush—which provides fuel for a fire. Another important factor is topography, or slope, which influences the speed of wildfire spread. Finally, access—or specifically lack of access—is a critical factor. We’ve learned through experience that when properties are on dead-end streets or in single-access neighborhoods, firefighters face added challenges.

In addition to examining those risk factors, our analysis also considered time-dependent, weather-related factors. California experienced extreme drought conditions this past summer, with hardly any rainfall, and we found that to be significant. We also looked at wind patterns.

Even before the Valley and Butte fires erupted, our assessment had placed a large number of properties destroyed or damaged in their wake at high risk based on location and vegetation risk factors. Our modeling also established the Valley and Butte locations as areas with high loss potential before the fires, as illustrated by the map to the right.

As illustrated on the maps below, in both fires we found a significant correlation between burn severity and the proximity of vegetative fuel. As we see more frequently, people are building homes and other structures in the California foothills, former wildlands dominated by vegetation. In such an environment, or wildland-urban interface (or intermix, depending upon proximity to wildland fuels), properties are at a higher risk for loss.

A good number of properties destroyed or damaged in the Valley and Butte fires were located on steep slopes—and steeper slopes can increase the speed and intensity of wildfire. Winds also played a role. We detected anomalous wind patterns, with wind gusts in the area of 30 miles per hour, in both the Valley and Butte fires. Those winds had a role in spreading the fires and making it difficult for firefighters to control the flames. The Valley and Butte wind patterns, however, didn’t approach the magnitude of the infamous 60-mile-per-hour Santa Ana winds, responsible for fanning California’s most devastating wildfires.

Wildfire trends

Are wildfires on the rise in California, and if so, are they growing in severity? Traditionally, the wildfire season in California has been summer and fall. But there’s growing evidence to indicate that it’s now becoming a year-round season.

Valley Fire: Burn Severity

FireLine Valley Fire, California map
Source: Maps and analysis provided by David Baryudin, Dominic Ford,
Hans Mayer, and Ting Wu of Verisk Insurance Solutions.

Butte Fire: Burn Severity

FireLine Butte Fire, California map
 
Home located on steep slope by Valley Fire
This home, located on a steep slope, was destroyed in the Valley fire.

However, the Valley and Butte fires don’t signal an increase in the severity of loss-causing fires. In fact, in terms of loss-causing fires occurring in California over the past 25 years, the Valley fire probably won’t break the top five when insured losses are adjusted for inflation. For example, if trended in today’s dollars, the total loss costs for the 1991 Oakland Hills fire—the greatest loss-causing fire in California’s history—would be about $4 billion. Other fires that may, and most likely will, surpass the Valley fire in severity, when losses are trended to the present day, include the Santa Barbara fire of 1990, the Laguna Beach fire of 1993, the Malibu/“Old Topanga” fire of 1993, the Cedar fire of 2003, the Old fire of 2003, and the Witch fire of 2007.

What does the future hold?

Lastly, when can we expect another fire like this year’s Valley fire? An evaluation of historical data suggests that a wildfire causing as much or greater loss as the Valley fire is likely to occur on average every three years or so. That’s not a new trend but a pattern that’s consistent with recent fire history. Since 1990, seven or eight such fires have occurred. Fortunately, the data also suggests that a fire causing losses on the magnitude of the Oakland Hills fire will be a much rarer occurrence.

Tomas Girnius, Ph.D.

Tomas Girnius, Ph.D., is a principal scientist with the research and modeling department at AIR Worldwide, a Verisk Analytics business. He previously led development of multitiered applications at Thomson Financial and was a computer specialist at the Smithsonian Astrophysical Observatory, where his work on the Chandra X-Ray Center (CXC) earned him the NASA Group Achievement Award.

Arindam Samanta, Ph.D.

Arindam Samanta, Ph.D., senior manager of underwriting products and analytics for Verisk Insurance Solutions, specializes in product and service development and management as well as scientific research. He manages Hail Damage Score™, Hail Risk Score™, and FireLine®, Verisk’s wildfire risk management product.


Resources:

Neil SpectorWebinar: An Analysis of the 2015 California Wildfires

Neil SpectorLearn about Verisk’s FireLine wildfire risk management tool

Neil SpectorRead our analysis of the 2015 California wildfires

Neil SpectorLearn about the AIR wildfire model

Neil SpectorDownload the 2015 Verisk Wildfire Risk Analysis

2015 California Fires

2015 California fires
*Damage assessment teams continue to collect information in the affected fire areas. Structures refer to homes, outbuildings (such as garages and sheds), and other structure types.

What puts one house at greater risk than another?

As a wildfire rages through an area, many homes can burn to the ground. Yet here and there stands an isolated house that escaped completely unscathed. Why? It’s a question that emerges after every wildfire event we’ve observed and analyzed.

Why are some properties spared? Is it just Mother Nature’s capriciousness? Or were the fortunate homeowners on to something?

It’s probably a little of both.

As a point of fact, the scenario of one home being spared while many others are destroyed is common. There’s practically no historical context for a major wildfire leaving a “scorched earth” landscape in its wake. A rare exception was California’s 1991 Oakland Hills wildfire, the worst in the state’s history. In that fire, every single property within the blaze’s footprint was destroyed. But that’s very uncommon.

To a degree, it’s just the luck of the draw—some homeowners are simply more fortunate than others. But some homeowners do in fact take measures that make their homes less susceptible to a wildfire’s wrath. Homes might have slightly thicker windows or better patched openings that leave no room for embers to penetrate. Or they may be set back farther from vegetation than neighbors’ homes, depriving the fire of necessary fuel.

What are some of the things that might give property owners in a wildfire risk area at least a slight advantage?

Start with the roof

Most houses destroyed during a wildfire are ignited by brands
or embers landing on the roof, so roofing materials matter. Wood shingle roofs are extremely vulnerable. Wood isn’t a protector;
it’s a source of fuel. Clay barrel tile roofs may be the best possible
roofing material. Stone-covered steel tiles are also extremely
fire-resistant. The new generation of asphalt shingles is excellent, although asphalt shingles will decompose more rapidly over time than other roofing materials.

The material used for siding is another important consideration. Siding materials respond differently to direct contact with flames and radiant heat exposure. Wood, of course, offers the least

Impact of building materials on fire risk
Why do some homes escape unscathed during a wildfire?

protection. Stucco is more fire-resistant than wood. Brick masonry is very fire-resistant. But here’s the rub: Masonry structures, while the most resistant to wildfires, tend to be brittle and are, unfortunately, vulnerable during earthquakes. In a state such as California, susceptible to both wildfires and earthquakes, this balancing act can limit a homeowner’s choices.

Regardless of the type of building material used, it’s important to make sure that the roof, frame, and siding are properly installed so that no cavities or openings exist. Fire embers can penetrate a structure through uncapped clay tiles, openings in wood soffits, and gaps in the siding.

Mind the brush

The primary factor influencing property risk during a wildfire is vegetation surrounding a structure and its proximity to the structure. Grass, trees, scrub, and brush are fuel for a fire. The greater the setback between a home and any vegetation, the better. We used to consider a 100-foot setback an optimal distance, but some scientists now believe that 30 feet is the critical threshold, with negligible benefit for farther setbacks.

Don’t forget Mother Nature

Mother Nature can be fickle, and luck has a lot to do with a homeowner’s chances of emerging unscathed from a wildfire. But better construction materials and smarter choices in siting a structure away from vegetation do provide a slight edge.