When 11,000 lightning strikes hit the San Francisco Bay Area on August 16, they ignited two of the largest fires in California history. Aclima had the pollution measurement devices in place to record the effects on local air quality. The company said the results show millions of Californians are breathing bad air, including many who may not realize it.

Twelve days later, three fires are still burning, and Aclima scientists have examined the impact of these lightning complex fires on air quality. The result is perhaps the most scientific analysis of data from a big fire in modern times (see the video above). For its analysis of the Bay Area fires, Aclima analyzed both its own data and the data collected by regulatory agencies and reported to the Environmental Protection Agency (EPA).

Aclima can measure air quality on a “hyperlocal” level using a fleet of electric cars with pollution-measuring sensors. The company gathers a massive amount of data compared to other pollution measurement collectors through this mobile method. Aclima previously used this data to assess the pandemic’s effect on car travel and pollution levels in San Diego.

Since the fires began, California’s inland counties have on some days experienced worse sustained daily air quality than Bay Area counties. But Bay Area counties saw larger maximum levels or spikes before the wind dispersed the smoke plumes and blew them inland. This has been tough for me, as I’ve been out jogging almost every day. Of the 168 days of lockdown, I have jogged for 159 days. For seven more of those days I have been writing an indoor exercise bike, and most of those due to the smoke.

Nights are better for walks

When looking at diurnal or day-to-night hourly patterns, Bay Area counties experienced worsening daytime and improving nighttime air quality, on average, from August 16 to August 25. That means it’s better to take the dog for a walk at night or in the early morning, said Aclima chief scientist Melissa Lunden in an interview with VentureBeat.

“I was seeing that by four or five o’clock that the levels were falling to a level where I could open the windows and cool the house down,” said Lunden, who has her own measurement device in her home. “We also have regular afternoon winds, and that blows it all inland.”

For this analysis, Aclima’s scientists focused on the daily average levels of fine particulate matter (PM2.5), which is a harmful pollutant at least 50 times smaller than a grain of sand and typically invisible to the eye. Even if you can’t see or smell smoke, you may be breathing air with unhealthy levels of particles generated by the wildfires.

Above: The air quality in the Bay Area changes during the day.

Image Credit: Aclima

To produce the embedded video, Aclima scientists analyzed regulatory air quality data from the state’s stationary monitors positioned at sea level throughout California and calculated changing daily average levels of fine particulate matter on a county-by-county basis throughout the state from August 15-25. The scientists then overlaid satellite-detected VIIRS fire hotspot data from NASA’s Fire Information for Resource Management System (FIRMS) to illustrate the changing air quality in relation to the locations of fire hotspots as seen from space.

As you can see, on many days the wildfires appear to impact daily average PM2.5 levels in inland counties more than Bay Area counties as the wind blew the smoke well beyond the fires.

In addition to analyzing daily air quality in California following the lightning complex fires, Aclima scientists analyzed data generated from the company’s mobile sensor network, which measures air pollution and greenhouse gases block by block across the Bay Area, day and night, weekdays and weekends.

Above: On August 22, the daily average PM2.5 levels in the Bay Area were lower than inland, but the daily maximums were at least as high in the Bay Area as inland.

Image Credit: Aclima

In Bay Area counties, a diurnal or day-to-night pattern showed cleaner air at the ground level in the evenings and early mornings, with the highest levels of PM2.5 at midday. Unlike regulatory monitors, the Aclima mobile sensor network takes measurements at various elevations on all publicly accessible streets.

Why is air better at night?

Lunden said on summer evenings an inversion layer of cool marine air is trapped beneath a layer of warmer air. The marine layer is stable, meaning there is no exchange of air between this lower level and the air above it. As the sun comes out and heats the ground, the height of this layer increases and there is more mixing of air both far above and near ground level. The difference in atmospheric pressure between the cool Pacific and the heating inland regions results in an onshore wind that starts to build mid-morning to a strong flow by late afternoon. As the sun sets, the evening inversion layer forms again.

Emissions from the fires are often found at higher elevations in the hills, and smoke rises high into the atmosphere. In the evening, this smoky layer is above the inversion layer and does not descend to the ground level. As this boundary layer grows during the morning, however, the smoke that has been held high above the ground mixes into this layer and the concentrations on the ground increase. Direct smoke emissions are also more likely to mix into this layer during the day. As the winds pick up, this smoky air is blown out of the Bay Area toward the east. And as the sun goes down, the reformation of the evening inversion layer results in clearer air being closer to the ground and where we breathe.

This isn’t to say that air quality at night has been good, or better everywhere, but it has shown a strong tendency to be measurably better on a county level throughout the Bay Area. For communities directly impacted by fire, unhealthy levels of smoke — not to mention the danger from the fire itself — have occurred at any time of day or night. And it’s important to note that these observed patterns hold true for what happened, but if the winds change then the patterns will too. A good resource for air quality is here.

People have tragically lost their lives to these historic fires, and many others have lost their homes. That’s not to mention the animals and habitats lost and the tens of thousands of people displaced due to evacuations. Meanwhile, millions of people are being exposed to unhealthy air quality across and well beyond California, Aclima said.

Air pollution is not confined to county, state, or country borders, and it is harming human and planetary health everywhere. By better understanding the impacts of climate change events — like lightning complex fires — Aclima said we can make informed decisions to protect ourselves and build a more resilient and equitable future.

Measurement challenges

You can check the health level of the air by zip code using Air Now, but you may not be able to entirely trust that estimate. That’s because the data is based on the government’s regulatory sensors operating in stationary places, and that data is then extrapolated to cover a much larger region. It’s not based on the fine-grained data Aclima collects with its cars, but it’s pretty much the best measure available at the moment.

Lunden said you can’t judge whether the air is safe enough to take a walk based on what you smell.

“What you smell from a fire is like the organic olfactory compounds that get emitted, and that stuff is pretty reactive in the atmosphere and disappears after some number of hours, but the smoke is still there,” Lunden said. “It just doesn’t smell like smoke anymore. It could still have pretty high concentrations.”

On top of that, in the Bay Area you can’t judge air quality by how blue the sky is. The inversion layer may or may not be in place, and you can’t see it. In other words, there isn’t a perfect way to know whether your air is clean or not. As for Aclima’s data from its cars, it isn’t analyzed in real time at the moment, so the company has to use it to analyze long-term trends, not the hourly changes in pollution levels you would need in order to understand whether it’s safe to go outside.

“The real strength and power in our data comes from the persistent differences we see,” Lunden said. “And those persistent differences come from repeat measurements over time. So we’ll be in your zip code on any given day, and then we’ll be there another two weeks later, and so on. And as we continue to do that mapping, you get those persistent differences.”


The audio problem: Learn how new cloud-based API solutions are solving imperfect, frustrating audio in video conferences. Access here