In my last post, I described our firsthand experience of flooding in Venice, and pondered the likely impact of climate change on Venice in the future. But that wasn’t our only firsthand experience of the impacts of climate change on our travels this summer. Having visited NCAR in July this year, we decided to come back to Boulder for the rest of the fall, to give me a chance to do more followup interviews with the NCAR folks, while I write up the findings from my studies of the software development processes for climate models.
Back in August I found a great house for us to rent, up in the mountains at Gold Hill. Shortly after I paid the deposit, I discovered the house was right in the middle of one of the most devastating forest fires in Colorado’s history. The fire, now known as the Fourmile Canyon fire, started on September 6, 2010, burned for over a week, affecting 6,181 acres, and destroying 169 homes. In terms of acreage, it wasn’t the biggest fire ever in Colorado, but in terms of destruction of property and damage costs, it was the worst ever.
I first heard about the fire while I was attending the Surface Temperature Record workshop in Exeter in September, and only then because of a conversation at dinner with some of the NCAR folks whose homes were in the evacuation zone. We spent the next few days wondering whether we’d have somewhere to live after all this fall, and trying to trace the path of the fire on various collaborative maps created by those on the scene. Not that we were affected anywhere near as much as the people who were evacuated, many of whom lost their homes and everything in them. But it gave us a taste of the impact of these massive forest fires on the communities who are affected.
Amazingly, the house we’re renting survived, even though several of the neighbouring houses burned down. Indeed, it seems amazing just how random the fire was – several patches of ground a few hundred yards from our house have been burned, but almost everything we can see from the house is untouched. The satellite images seemed to show huge areas completely devastated, but in reality, the affected area is now a real patchwork of healthy trees and burned sections.
But this patchwork effect is actually easy to understand once you build a good computational model. I particularly like this NCAR simulation of the spread of forest fire. Notice how the prevailing winds (shown by the arrows) push the fire forward, but also how the updraft from the fire affects the wind pattern to the sides and in the path of the fire, effectively funnelling it into a narrower and narrower path. This certainly corresponds to the stripes of fire damage now visible in the area of the Fourmile Canyon fire, and explains why the fire damage seems so patchy.
As this fire was unusually large by Colorado standards, I wondered about the impact of climate change. In particular, I thought the damage caused by Mountain Pine Beetles might be to blame. When we drove up to Breckenridge in July, the kids noticed that many of the trees were dead, and we googled a little back then to discover it was because the hotter, drier summers were encouraging the spread of the pine beetles, and weakening the trees’ defences. And from a major study published in September this year, it’s clear that climate change is a major factor, and the destruction to pine forests across the North American Rockies will only get worse as climate change progresses.
So I figured all those dead trees would just encourage bigger wildfires. But, as usual with climate change, it’s not that simple. In particular, the areas damaged by fire this year don’t correlate with the areas most damaged by the beetles. It looks like the trees killed by beetles are actually less susceptible to fire, because the needles drop to the forest floor and decompose fairly quickly, while the trees lose the oils that encourage fire in the tree canopy. But although the beetle damage doesn’t cause the fires, climate change affects both, because the hotter drier summers increases both the spread of the beetles and the likelihood of fires.