The importance of active forest management
This tweet below immediately caught my attention. A green piece of land in the middle of the Bootleg fire.
The image suggests that the green area was far less damaged than the other areas surrounding it because it was treated with prescribed fire. The data scientist in me couldn’t resist to have a closer look at this area. Luckily, with satellite imagery we can.
By aggressively suppressing wildfires for decades we’ve seen an accumulation of fuel loads. Active forest management and prescribed burning are potential solutions to reduce the risks of megafires.
For centuries, tribes have been using small intentional burns to reduce the risk of larger, more dangerous wild fires. In the area of the Bootleg Fire, the Klamath Tribes worked with the US Forest Service and other cooperators to thin young trees and apply prescribed burning.
What makes the image interesting is the difference in damage between the thinned forest and the thinned forest that has also been treated with prescribed burns. First, we want to confirm that both areas have similar characteristics.
Pre-fire very high resolution satellite imagery confirm that the area where thinning has been applied show similar characteristics as the area where thinning and prescribed burns have been used to reduce the fuel load. Also other characteristics, like the slope of the area are similar.
There is a clear difference with areas surrounding the thinned areas, where we can see more dense vegetation.
So what did cause the difference in damage? By looking even further back in time, on satellite imagery we can see that the areas that have been least impacted by the Bootleg Fire were burning in April of that same year. This must be the prescribed burns because the footprint looks controlled and match the boundaries of the forest plot without affecting the surrounding forests.
The effect of the prescribed burns is best captured by comparing the pre-fire, after prescribed burning, image directly with the post-fire image. In this false colour composite healthy vegetation shows up in red. On the image from April 2021 we can see that the area has been treated with prescribed burns, burning away low vegetation and clearing out fire-fueling vegetation. At the same time, we can still see healthy trees in red in the same area. Also note, the healthy vegetation outside the area that has been treated with prescribed burns — a lot of healthy trees.
The difference with the post-fire image couldn’t be bigger. The Bootleg fire burned large areas of land in just a few days. Fully consuming most trees, and full scorch or high scorch for the area that has been thinned — almost no “red” trees left. While the area that has been thinned and treated with prescribed burns mostly show low scorch and unburned trees — mostly “red” trees.
Obviously, more data is needed to draw any conclusions. I strongly believe that more accurate, up-to-date data can help improve decision making. Satellite imagery and machine learning can play an important role, but only if combined with local on the ground knowledge. Collaboration is key to help protect our forests.
Edit: Bill Tripp made a good point that “No Treatment” in the original photo is probably not fair. There has been some activity in the area for the last 20 years. By using historical satellite imagery we can see that the “No Treatment” area has partly been thinned (yellow area) and clear cut (red area) in 2003. The areas “Thinning + Prescribed Fire” and “Thinning Only” seem to have been treated in 2017 (green area).