November 9, 2019
How Bad Science & Horrific Journalism Misrepresent Wildfires and Climate
As one wildfire expert wrote, “Predicting future fire regimes is not rocket science; it is far more complicated than that.” But regardless of accuracy, most people are attracted to very simple narratives such as: more CO2 causes global warming causes more fires. Accordingly in the summer of 2019, CNN trumpeted the headline California wildfires burn 500% more land because of climate change. They claimed, “the cause of the increase is simple. Hotter temperatures cause drier land, which causes a parched atmosphere.” CNN based their claims on a scientific paper by lead authors Park Williams and John Abatzoglou titled Observed Impacts of Anthropogenic Climate Change on Wildfire in California. The authors are very knowledgeable but appear to have hitched their fame and fortune to pushing a very simple claim that climate change drives bigger wildfires. As will be seen, their advocacy appears to have caused them to stray from objective scientific analyses.
If Williams and Abatzoglou were not so focused on forcing a global warming connection, they would have at least raised the question, ‘why did much bigger fires happen during cooler decades?’ The 1825 New Brunswick fire burned 3,000,000 acres. In Idaho and Montana the Great Fire of 1910 burnt another 3,000,000 acres. In 1871, the Great Michigan Fire burned 2,500,000 acres. Those fires were not only 6 times larger than California’s biggest fire, they occurred in moister regions, regions that don’t experience California’s Mediterranean climate with its guaranteed months of drought each and every summer. If those huge devastating fires occurred in much cooler times, what are the other driving factors of big wildfires?
Bad analyses cause bad remedies, and here is why Williams and Abatzoglou’s last paper exemplifies a bad scientific analysis. Analyzing changes in California’s burned areas from 1972 to 2018 they claimed, “The clearest link between California wildfires and anthropogenic climate change thus far, has been via warming-driven increases in atmospheric aridity, which works to dry fuels and promote summer forest fire.” But natural cycles of low rainfall due to La Niñas also cause dry fuels. The increase in burned area is also attributed to increases in human ignitions such as faulty electrical grids, to increased surface fuels from years of fire suppression, and to changes in vegetation that increased the abundance of easily ignited fine fuels like annual grasses. Furthermore, temperatures in some local regions experiencing the biggest fires have not been warming over the past 50 years (See temperature graphs in this essay’s last segment. Data from Western Regional Climate Center). All those factors promote rapid wildfire spread and greater burned areas. Although good science demands separating those contributing factors before analyzing a possible correlation between temperature and area burned, Williams and Abatzoglou oddly did not do so! That’s bad science.
Although Williams and Abatzoglou did acknowledge that other factors modulate the effects of warming on burned areas they admitted their statistical correlations did not “control” for those effects. To “control” for all those contributing factors, they could have easily subtracted estimates of burned areas associated with those factors. For example, a 2018 research paper estimates, “Since the year 2000 there’ve been a half-million acres burned due to powerline-ignited fires, which is five times more than we saw in the previous 20 years.” Did Williams and Abatzoglou not do the needed subtractions of other well-established factors because it would weaken their global warming correlation?
Similarly, CNN journalists were content to simply blame climate change. However, in light of the increasing devastation caused by powerline-ignited fires, good investigative journalists should have asked the former California Governor Jerry Brown if he now regrets having vetoed the bipartisan bill crafted to secure the power grid; an action that could have saved so many lives and property. Instead CNN simply promoted Brown’s persistent climate fearmongering quoting, “This is only a taste of the horror and terror that will occur in decades.”
Ignoring the complex effects of human ignitions, CNN also parroted claims that global warming is causing fire season to last all year. But as seen in the graph below from a 2017 wildfire study, the United States’ natural fire season is due to lightning and only dominates during the months of July and August, when California’s high wind events are low. In contrast it is human ignitions that extend fire season, dramatically increasing ignitions throughout the winter months when fuel moisture is higher, and into seasons when cooling desert air generates strong episodes of Santa Ana and Diablo winds. Those high winds cause fires to spread rapidly, burning 2-3 times more area than fires ignited during low winds, and California’s most destructive fires recently occurred during those high wind events. However, like other researchers, Williams and Abatzoglou reported no trend in those destructive California winds. Furthermore, climate models suggest a warming climate should cause weaker winds. So, without a change in California’s windy conditions, high winds can’t be blamed, directly, for the increased burned areas. However, because more human-caused ignitions occur during the winter, it increases the probability that more fires will be amplified by those strong winter winds. As US Geological Survey’s wildfire expert states, “Some will argue that it’s climate change but there is no evidence that it is. It’s the fact that somebody ignites a fire during an extreme [wind] event.”
The timing of human ignitions is but one driver of more and bigger fires. Increased surface fuels are another huge factor. It is well known that past fire suppression has allowed surface fuels to accumulate in forests, leading to bigger and more devastating fires. But the changes in surface fuels are more complex. Some scientists point out that certain logging practices spread “invasive grasses called cheat grass, for example, and other ones that form this really thick mat across the area after logging and that grass just spreads flames very rapidly and fires burn very intensely through that.” California’s Democrat congressman Ro Khanna has been arguing that the U.S. Forest Service policy to clear cut after a wildfire is making California’s forest fires spread faster and burn hotter by increasing the forest floor’s flammable debris. Khanna says, “Because we don’t have the right science, it is costing us lives, and that is the urgency of getting this right.”
Controlling the spread of cheat grass is urgently needed. Grasses are “fine fuels” that ignite most easily. The 2018 Carr Fire was California’s 7th largest fire and threatened the town of Redding, California. It started when a towed trailer blew a tire causing its wheel rim to scrape the asphalt creating a spark which ignited roadside grasses. Those grasses carried the fire into the shrublands and forests. Grasses are classified as 1-hour fine fuels, meaning they become highly flammable in just one hour of warm dry conditions. Climate change is totally irrelevant. It does not matter if it was wet and cool, or hot and dry during previous days, weeks or years. Just one hour of warm dry fire weather sets the stage for an explosive grass fire that then gets carried into the forests. Fire weather happens every year, and partially explains why fires could burn 3,000,000 acres in the cool 1800s.
It was not human ignition but lightning that caused the 2012 Rush Fire. It was California’s 4th largest fire burning 272,000acres of sagebrush habitat, which then continued to burn additional area in Nevada. Historically, because surface fuels are scarce, hot dry sagebrush habitat rarely burned (once every 60-100 years). But invasions of non-native cheat grass have now provided ample fuel to turn small lighting fires into huge conflagrations. Eleven of the USA’s 50 biggest fires in last 20 years are in the Great Basin, where invasive cheatgrass is spreading. Nevada’s largest fire was the 2018 Martin Fire. Rapidly spreading through the cheat grass, it burned 439,000 acres. Cheat grass fires are a great concern for biologists trying to protect the threatened Sage Grouse as cheat grass-dominated sagebrush habitat now burns every 3-5 years. Habitat with high cheat grass abundance are “twice as likely to burn as those with low abundance, and four times more likely to burn multiple times between 2000-2015.”
When experts estimate impending fire danger, they determine how fast a fire will spread. The Spread Component considers the effects of wind and slope and daily changes in the moisture content of the surface fuels. Large dead trees may become flammable after 1000 hours of warm dry conditions, but still thick fuels only ignite if fast burning surface fuels supply enough heat. Thus, the Spread Component only considers smaller-diameter fuels like grasses that can dry out in an hour, as well as twigs and small branches that dry out within 10 to 100 hours. Central and Southern California are dominated by shrubby habitat with small diameter fuels that allow fire to spread rapidly. The December 2017 Thomas Fire was California’s 2nd largest fire. Its human ignition coincided with a Santa Ana wind event resulting in the burning of 282,000 acres in southern California.
Counter-intuitively Williams and Abatzoglou found the correlation between burned area in the hotter and drier climate of California’s Central and South Coast to be “relatively weak”. Accordingly, they reported “Annual burned area did not change significantly in Central and South Coast.” That insignificant climate effect over half of California escaped the notice of journalists who only cherry-picked the researcher’s more alarming climate narratives. Most interesting, Williams and Abatzoglou suggested the lack of a climate-change correlation with California’s Central and South Coast burned areas was because fires there were “strongly manipulated by humans via ignitions, suppression, and land cover change.”
Lightning is rare along California’s Central and South Coast, so nearly 100% of those fires are ignited by humans. As California’s population doubled since the 1970s, adding 20 million people, the probability of more human-started fires has increased. Unlike forested areas where fire suppression builds up deadly surfaced fuels, California’s Central and South Coast need to suppress fires. Due to more frequent fires caused by humans, shrublands are converting to grasslands. The increased fine fuels of the grasslands more readily ignite and spread fire. Furthermore, California’s natural climate undergoes wet years due to El Nino followed by dry La Nina years. Wet years make fine fuels more abundant. Thus fire suppression is needed to prevent more frequent fires caused by the conversion of shrublands to grasslands.
In contrast to the insignificant changes in burned areas in California’s southern half, Williams and Abatzoglou reported burned areas in the Sierra Nevada and the North Coast increased by more than 600%, which they attributed to human-caused climate change. They reported, “During 1896–2018, March–October Tmax [maximum temperature] averaged across the four California study regions increased by 1.81 °C, with a corresponding increase in VPD [ Vapor Pressure Deficit - a measure of atmospheric dryness] of 1.59 hPa (+13%)…The observed trends in Tmax and VPD are consistent with trends simulated by climate models as part of the CMIP5 experiments, supporting the interpretation that observed increases in California warm-season temperature and VPD have been largely or entirely driven by anthropogenic forcing.”
But how can only half of California’s fires be due to global warming and the other half not? All of California is “strongly manipulated by humans via ignitions, suppression, and land cover change”? Were Williams and Abatzoglou straying from objective science?
Part of the problem is their ill-advised use of a maximum temperature averaged for all California. Several studies have reported that maximum temperatures in the northern half of California have not exceeded the high temperatures of the 1930s. Because the early 20th century temperatures were deemed natural, unless recent temperatures exceed the natural 1930s, then human-caused warming is unlikely. Curiouser and curiouser, southern California has experienced temperatures that exceeded the 1930s. Yet there Williams and Abatzoglou did not find a significant effect from climate change.
Regardless Williams and Abatzoglou claimed “The clearest link between California wildfire and anthropogenic climate change thus far, has been via warming-driven increases in atmospheric aridity, which works to dry fuels and promote summer forest fire.” Yet summer maximum temperatures, averaged from March through October, located in the vicinity of California’s big fires do not indicate global warming. For example, the August 2013 Rim Fire centered around Yosemite National Park, was California’s 5th largest fire and 2nd largest in northern California, burning 257,000 acres. It was started by a hunter’s illegal campfire that he let get away. Unfortunately, there is no cure for stupid. Nonetheless, Yosemite’s maximum temperatures were warmer in the early 1900s. However, an in depth study of the Rim Fire found a strong correlation with the amount of fast burning shrubland interspersed with the trees.
The November 2018 Camp Fire was California’s deadliest fire destroying the town of Paradise. It was also its 16th largest fire burning 153,000 acres. It was ignited by a faulty power grid during a strong Diablo wind event. Similarly, based on weather data from nearby Chico CA, maximum temperatures were higher in the 1930s.
The Mendocino Complex Fire was California’s largest fire (since 1932). In July of 2018 it burned 459,000 acres. The source of human ignitions is still under investigation. Still, those fires were centered around the town of Ukiah which also reveals a cooling trend since 1930.
In the October 2017 wine country fires, the Tubbs Fire was the 4th deadliest. It only burned 37,000 acres but high winds drove embers into the dwellings of the heavily populated outskirts of Santa Rosa. Again, global warming was irrelevant as Santa Rosa has experienced a cooling trend since the 1930s.
Still some people are determined to link catastrophic fires with climate change. So, they will suggest delayed autumn rains allow more late season ignitions or the fall fires to burn longer. In Williams and Abatzoglou’s abstract they claim, “In fall, wind events and delayed onset of winter precipitation are the dominant promoters of wildfire.” But their results found, “no all-region trend in onset of winter precipitation or October–November wet-day frequency during 1915–2018.” As illustrated below by the October precipitation data for Santa Rosa, since 1900 there’s a 10% chance no rains will fall in October. Furthermore, October experienced more zero rainfall months in the early 1900s. A global warming caused delay in autumn rains has not yet been detected.
So, doing my best Greta Thunberg imitation, I say to climate alarmists, “How dare you misrepresent the causes of wildfires. How dare you imply less CO2 will reduce human ignitions and reduce surface fuels and the spread invasive grasses. Bad analyses lead to bad remedies! Your bad science is stealing Californian’s dreams and your false remedies distract us of from the real solutions. Young people and old alike, must demand better science and better journalism!”
Jim Steele is Director emeritus of San Francisco State’s Sierra Nevada Field Campus and authored Landscapes and Cycles: An Environmentalist’s Journey to Climate Skepticism