California fires: Tracking the smoke
Two years ago, in August 2018, we wrote about California’s record-breaking fire season that saw more than 11 severe blazes burn a combined total of 820,000 acres. By the end of the year, 1,975,086 acres had burned.
Devastatingly, 2020 has beaten 2018’s record by almost twice as much. According to The Department of Forestry and Fire Protection, 4,105,786 acres have burned since January 1st 2020. However this time, the fires have caused less damage to buildings — destroying 9247 structures compared to 24226 in 2018. This is probably due to the location of the fires and let’s hope it doesn’t increase.
The image below, using data from Cal Fire, visualizes the amount of acres that have burned each year since 2013. This bar chart shows the dramatic escalation of wildfire intensity over the last seven years. To find out how we made this graph, visit the notebook.
Gigafire — A New Phenomenon
The August Complex Fire that started as 38 separate fires caused by lightning strikes, has now developed from a “megafire” (burning over 100,000 acres) to the first “gigafire” (burning over a 1,00,000,000 acres) in the history of modern California. This fire is larger than the state of Rhode Island. Australia’s bushfire earlier this year reached gigafire status, but California’s is the first gigafire in the United States for ten years.
High-Resolution Rapid Refresh (HRRR) Smoke Model — CONUS 3km — Experimental: Now available in Datahub
To better understand the extent of the wildfires in the Western United States, the High-Resolution Rapid Refresh Smoke (HRRR Smoke) model can be used. This is a three-dimensional model that allows simulation of mesoscale flows and smoke dispersion over complex terrain. The smoke model comprises a suite of fire and environmental products that can be used for forecasts during the wildfire season. The source of this experimental dataset is Global Systems Laboratory. While the dataset is useful for modeling fire smoke forecasts and fire intensity, it is an “experimental” dataset meaning that there are some limitations including missed detections and model inaccuracy. Therefore, this data should not be used to make decisions regarding human safety or property. This smoke product will become available in the Operational HRRR in the December timeframe; as part of the next upgrade.
Applying the Data
In light of these fires and our available datasets, we analysed the smoke from the wildfires on the lowest model level (8 meters) across the entire country and at the Californian county level. For a tutorial on how to do this, the GitHub notebook is available here.
Visualizing the Smoke: United States
As you can see from the visual below, smoke travels far and is erratic, changing direction in a matter of hours. This forecast is from 4th October 2020 and it clearly shows the immense spread of smoke across California and the rate at which this smoke spread to neighboring states.
The second animation below details smoke from the 14th October 2020. We can see that there is a lot less smoke than there was on the 4th, however the smoke has traveled far, having been directed by changing weather patterns.
Closer look: California
As most of the fires are spreading throughout California, it makes sense to take a closer look at what is happening across the state. The HRRR Smoke model has a good spatial resolution at 3 km, which makes tracking the smoke on a more granular level much easier.
Below we can see that the smoke covered most of the state last week. However, the more recent forecast shows that the situation has significantly improved this week.
From this week’s forecast, pictured below, we can see that the smoke is much more localized. While it is not traveling as far, the smoke still presents a severe threat to the health of humans and our ecosystems.
Health issues: Smoke leads to poor air quality
Smoke from wildfires contains thousands of individual compounds, including carbon monoxide, volatile organic compounds (VOCs), carbon dioxide, hydrocarbons and nitrogen oxides. The most prevalent pollutant by mass is PM2.5, which refers to atmospheric particulate matter (PM) that has a diameter of less than 2.5 micrometers. A product of combustion, these fine particles tend to stay in the air longer than heavier particles, thus increasing the chances of inhalation by humans and animals. These minute particles can penetrate the lungs and circulatory system, often causing adverse health effects.
Luke Montrose, the environmental toxicologist, explains how those PM2.5 particles affects human body:
The human body is equipped with natural defense mechanisms against particles bigger than PM2.5. As I tell my students, if you have ever coughed up phlegm or blown your nose after being around a campfire and discovered black or brown mucus in the tissue, you have witnessed these mechanisms firsthand.
The really small particles bypass these defenses and disturb the air sacks where oxygen crosses over into the blood. Fortunately, we have specialized immune cells present in the air sacks called macrophages. It’s their job to seek out foreign material and remove or destroy it. However, studies have shown that repeated exposure to elevated levels of wood smoke can suppress macrophages, leading to increases in lung inflammation.
For further information about air quality, our View2020 dashboard gives a short overview of how PM2.5 and other air quality parameters have changed in California over the course of the year.
What’s next: comparing HRRR Smoke with CAMS air quality forecast
In the next blogpost and Notebook we are planning to compare CAMS forecast particulate matter variables with HRRR Smoke forecast to see if there are any similarities. This comparison will be useful for anyone grappling with smoke forecast data.
We routinely add new datasets to Planet OS. If you’d like to be notified when new data becomes available, follow Planet OS on Medium or subscribe to our email newsletter to receive future updates in your inbox.
