Understanding your Air Quality numbers

 As our coastal sites turn green this morning and we start to see hope for clean air for the rest of us later today and tomorrow, I thought I’d take a minute to clarify some common misconceptions regarding air quality numbers, where they come from and what they mean to you.  The two indices, Air Quality Index (AQI) vs. Washington Air Quality Advisory (WAQA), have created a lot of confusion.

Both AQI and WAQA are a unitless index calculated from a given concentration of air pollution. There are several different measured pollutants that affect air quality: PM2.5, ozone, NOx, CO, and SO2. PM2.5 is the most commonly measured and generally the one of most concern in Washington state, especially during wildfire smoke episodes. Although all these pollutants can affect AQI and WAQA values, I’m only going to focus on PM2.5 in this post.

State and Federal government run air quality maps (https://airnow.gov and https://enviwa.ecology.wa.gov/home/map) use the same air quality monitors and data to determine air quality in your area, but they use a different index to represent what that air quality means for your health.

So, why are there two different indices?  Although we assign breakpoints to the different categories, air quality is a spectrum. Every person will be affected by poor air quality differently.  These are general guidelines and the breakpoints are based on statistical assessments of how large numbers of people respond to varying levels of poor air quality. EPA studies have assigned risk at a certain point, while Washington state toxicology research found more protective levels would better serve our community. Looking at the breakpoints in this table, a PM2.5 concentration equal to 50 ug/m³ is considered unhealthy by WAQA standards, but unhealthy for sensitive groups when using the AQI scale. You are not a statistic and may be adversely affected at moderate levels, or perhaps you can go hiking without difficulty when conditions are unhealthy for sensitive groups.  For individual health concerns, your best bet is to consult your doctor.

AQI and WAQA health risk categories are based on a 24-hour exposure. The number you see on the map is updated hourly and represents the previous hourly average weighted by the air quality observed over several hours before. For example, if the air quality is green all morning and a nearby building catches on fire, smoke may affect the monitor for one or two hours at unhealthy (red) levels, before returning to green. The monitor may only show unhealthy for sensitive groups (orange) during those two hours since the exposure time was so short. As we observed at the beginning of this smoke event, heavy smoke rolled into some areas very quickly and air quality went from good to hazardous in 30 minutes. The map will not update that value until the hour of data collection is complete and may take another hour or two to catch up to instantaneous air quality. Both airnow and the state maps have the option to look at the actual hourly PM2.5 concentration from the previous hour. Those numbers will be the same on both maps.

The second question we often get is “who is more trustworthy? My air quality app, a third-party air quality reporting site, airnow, or the state air quality map”? It is important to understand all these sites are using the same data, if they are using actual air quality measurements. We don’t often know how third-party apps and websites calculate the AQI or interpret our data. In that sense, we consider EPA and the state map to be the gold standard as we are directly involved in collecting the data, making sure it’s valid, and reporting the concentrations.

 

Since this is the era of big data, lets compare our plight with those from far and wide

There are claims that we've been exposed to the worst air in the world. True? Fair comparison? Let's investigate that without belittling the terrible air quality conditions we've all had to endure for at least a week now.  

Here's a map of global air quality right now. It uses all available measurements worldwide and uses machine learning methods for spatially interpolation. US west coast is in baaad shape. Worst ranked cities right now are all in our neck of the woods. 

But to do a proper apples-to-apples comparison, we should compare our air during our "bad" season with the "bad" seasons in other places, not our bad against their better times. So lets leverage data from the US State Department's (DoS) air quality monitors around the world. These are easily obtained from one place and have a pretty good quality control regime. 

I've pooled PM2.5 data from the ten highest embassies which recorded data for at least two years anywhere between 2012- 2019. These are: New Delhi & Kolkata (India), Dhaka (Bangladesh), Ulaanbaatar (Mongolia), Kampala (Uganda), Kathmandu (Nepal), Jakarta (Indonesia), Chengdu & Shenyang (China), and Manama (Bahrain). You can see more details about their air quality here. "Normal" conditions were defined as the range between monthly lower and upper quartiles, since the median passes right in the middle of it. This is also known as the interquartile range, or IQR.

The above figure shows how air quality in Seattle, Spokane, Vancouver, Omak and Yakima have varied from 1 Jan- 15 Sept 2020, against the backdrop of IQRs at ten embassies. As can be seen, the overseas cities experience terribly compromised air quality primarily in the winter months. Right now is their "better" season. Even though we spiked right to the top of the AQI chart this week, bear in mind that:

• All these overseas locations see concentrations that are higher than their own IQRs 25% of the time. Or put another way, they record even higher concentrations for 3 months of the year.
• Our air will not remain this bad for several months. Most WA sites have experienced some improvement already and we're still on track to clear out in the Friday- Saturday timeframe. 
• In spite of our wintertime temperature inversions and woodsmoke concerns, we're still in far better shape at that time of year.

Moral of the story: yes its bad here now. Many others have it worse for much longer. Folks who live in these environments and our diplomats serving there can offer us some coping tips. 

Where's the smoke coming from now?

One key thing that we need to get smoke out of the air is to stop putting smoke into the air. So where is the smoke coming from now? Here in Washington we have some active fires although the map of Washington below is somewhat misleading as only a few of the fires shown are producing significant amounts of smoke. Three fires in the state were listed as showing moderate fire behavior today including the Big Hollow, the Inchelium Complex, and the Cold Springs fires. The others including Whitney, Fish, Apple Acres, Pearl Hill, Sumner Grade, Customs Road, Babb, Manning Road, and Evans Canyon are listed with minimal fire behavior so they are not smoke free but are unlikely to be making significant contributions to our impaired air quality outside their local area.

But to the south in Oregon there are numerous very large fires that are actively burning and sending smoke north to Washington. The northern group of fires including Riverside, Beachie Creek, and Lionshead, plus a few small ones nearby have a combined size of nearly 500,000 acres and are burning in heavy timber, slash, and brush fuel types. But that’s not all, numerous other large fires stretch the length of the Oregon Cascades and into California. Recent wind conditions have been moving smoke from these fires to the north and right towards us. The smoke layer has actually resulted in cooling many of the fires somewhat although as the smoke begins to thin out somewhat the fires are expected to become more active. More smoke, less fire, less smoke, more fire? It sounds like an unfortunate cycle. 

But on the horizon, arriving Thursday and lasting until Saturday, there is a prediction for enough rain in the Oregon Cascades to tamp these fires down considerably. Will they go out? No. And the farther south in Oregon you go, the less the rain will result in significant suppression of the fires. But the hard work of the firefighters combined with an assist from the weather gods should result in a pretty decent reduction in smoke production by later this week. Let’s hope so!

Is this the worst smoke we've had?

My colleague Dr. Beth Friedman crunched through statewide fine particle pollution data and prepared these plots comparing this year's smoke events against the last 5 years. The y- axis of the plot below shows the % of time all sites in the state recorded air quality in each of the stated categories. Monitor downtime and operating schedules have been accounted for. Only levels corresponding to an exceedence of the federal standard are shown for clarity.  

Washingtonians have not spent as much time breathing compromised air this year compared to 2017 and 2018. We had a late start to the wildfire season this year. However the amount of time we have spent breathing Hazardous air is unprecedented. Even if we look as far back as 2006, we don't see Hazardous conditions occurring for anywhere as long. 

Next question is where those conditions occurred (you guessed it: everywhere). Beth mapped out how different areas in WA compared against each other, by averaging concentrations from the highest 3 days for each site this September. The size of the circles is number of days greater than the federal standard. 

Seems like southern Yakima county followed by parts of Okanogan County had the most number of polluted days this month, but the worst air was recorded in the Columbia River Gorge and Clark County. Northwest and southeast WA had the "cleanest" and fewest bad days, respectively. 

Take home message: Not the longest we've had to endure crummy air, but this is the dirtiest air we've had to breath as a state. And the season isn't over yet. 

Overnight drizzle didn't move the smoke needle much.

Slight welcome drizzle overnight in parts of western WA and a marginal improvement in fine particle pollution levels along the coast. Not much, but we'll take what we get. Air is still Unhealthy or worse in most of the state. 

All the smoke models that ran overnight overdid the amount of clearing this weather system brought. According to them there should have been areas of Green and Yellow on the map by this morning. How we wish they were right! Therefore we're we're going for a persistence forecast. That is, expect the status quo to continue. Not pinning too much hopes on the weather system and slight rain expected tonight either. Substantial clearing is will have to wait until the Friday- Saturday timeframe. Here's the forecast for tomorrow. Nothing we havent already seen. 

Later today we will provide some data analysis comparing this smoke season to past years. If our readers would like to see informative data products, please let us know and we will try, within reason, to prepare the most relevant plots/ tables. A friend of mine wanted me to issue a hyper-local, personalized forecast showing how his house will have clean air while everyone else had to endure hazardous conditions. I said no. Same goes for data products. 

Atmosphere is not delivering what we ordered

Sad to say, the clearing that should have been here by now is not only tardy but is poofing out as a lackluster weather feature that won't do much for us. The "cleanest" air in the state right now is Unhealthy air, whereas the model's advertised menu said we should have been seeing substantial improvements along the coast this morning. 

Updated projection

Not much in terms of weather to change the status quo a whole lot. See the forecast map for expected conditions in your area. To add to our woes, light southerly winds will continue for another day at least, dragging more smoke directly from Oregon fires northward along the I-5 corridor. So even if the ubiquitous smoke pool from offshore starts to erode a bit, a replacement is en route. 

Wanna get geeky and figure out what went wrong?

Many of us had relied on the HRRR- smoke model for several reasons. The University of Washington runs a suite of weather models which all have different physics options, so as to get a handle on uncertainty. One of the ensemble members is run with the same physics options as the HRRR. The plot below shows what yesterday's UW model runs thought ventilation indices along the WA coast would be like. The HRRR (light green trace) was in fact one of the most optimistic, showing about twice as much clearing as the mean of all other models. So today's forecast leans closer to the UW WRF model, which is closer to the ensemble mean.

Where are you when we need you most, O clean marine air?

"I'm already there. Take a look around. I'm the shadow on the ground". 

OK that doesn't help. But its true. Fog around parts of the Puget Sound confirm that marine air did get through the Chehalis gap overnight. The trouble is there's still plenty of smoke over the coastal waters which the marine air is dragging along with it. But this morning's satellite images show where the silver lining is, and how the vertical smoke column over WA is thinning. 

Ground level smoke is not going to erode as fast as we'd like, however. It will be Monday before much of western WA and the central and northcentral foothills of the Cascades see substantial relief. For far eastern WA, plan for little relief before Tuesday. I wish I had better news for areas around the Columbia River Gorge that will be directly downwind of some monstrous Oregon fires. 

Check the forecast map and plan accordingly. As has been the case since Friday, our best advice is to remain indoors and hunker down. 

Some of Ecology's monitors in eastern WA are not reporting data to the map above because the high readings are being automatically invalidated ("nah, that's erroneous data, no way it can be so high --> trash & blacklist!"). We've just convinced the algorithm to come to terms with the bitter truth and these data should start to appear on the above map shortly. Ecology's monitoring data dotmap has been current, however. 

An aside: I've received questions from folks who're wondering why Ecology's map shows Good air quality in some areas, when everything else nearby is crummy. Thats because the default view of the map shows all pollutants being monitored, not just fine particle pollution (PM2.5, aka "smoke"). Ozone and sulfur dioxide levels are low, so sites monitoring just these pollutants show up as Good. Click on the "PM2.5" tab to filter these out. 

Gasp! How much longer???

Quick answers: at least another day and a half in Western WA. 2-3 days in eastern WA.

Gradual clearing will commence on the WA coast on Sunday from west to east, and it will be Monday before that pushes across the state. For western WA, this means we're close to the peak of the episode, but much of eastern WA will deteriorate further today before it starts to get better. The size of the Oregon smoke plumes parked offshore is so "super-massive", and the fires themselves are very smoky, so smoke will continue to pour into the state for a while to come. And there are also several fires within WA to contend with.

We've received an overwhelming number of public queries on the blog from concerned citizens and kindly ask that folks don't take offense if we missed responding to you individually. 

Answers to most commonly raised questions

• There are no pockets of clean air to retreat to this weekend. Your favorite campground or hiking trail isn't going to be magically shielded from smoke, no matter what the elevation (except perhaps subsurface caves... but I digress). Here's the latest satellite picture- that's all smoke over the state, very little clouds. See what I mean? 

• Please use this forecast map to self- serve. Forecasts beyond two- three days are less certain
• Resources to protect your health during these events can be found here. Best advice at this time is stay indoors. 

Statewide Air Quality Alert Through Weekend

The Department of Ecology issued a statewide air quality alert yesterday, which continues through the weekend.  Smoke from many fires across the region is impacting our state.  Please keep in mind we are unable to answer the large volume of questions posted in the comments on this blog, but will do our best.  Forecast information for your region is available on the Smoke Forecast Map. The forecast map attempts to predict a 24-hour daily average, but only for locations that have a regional monitor.  Air quality monitor maps are the best source for assessing current conditions (e.g. the map at the top of this blog, EPA's AirNow page, or the Ecology WAQA page). With such a large area experiencing poor air quality, it's best for people to stay indoors.  See yesterday's blog post Wildfire Smoke & COVID-19 for information on how to protect your health.

Calmer winds and hazy skies helped to keep large fire growth moderated yesterday and allowed for fire-fighting progress in Washington.  However, the Big Hollow fire in Southwest Washington still has active fire behavior with no containment. 

Today's GOES image (below) shows that the extent of smoke currently covers most of Western Washington and parts of Central Washington.  Smoke is expected to continue its path across the state, impacting Eastern Washington later today.  Overall, air quality is expected to slowly start getting better, from West to East, on Sunday. 

    GOES Image (Sept 11, 9 am)

Wildfire Smoke & COVID-19: A Bad Combination for Health

The active wildfires in our state present many dangers, including the impact wildfire smoke has on our health. The smoke produced by wildfires can also be dangerous to you and your family, even when you don’t live near the wildfire.   

Breathing in wildfire smoke can cause symptoms that are relatively minor, such as eye, nose, and throat irritation, and also more dangerous symptoms like as wheezing, coughing, and shortness of breath. We are especially concerned this year with COVID-19 because both impact our respiratory and immune systems and some of the symptoms are the same, like coughing, wheezing and shortness of breath. If you have COVID-19, breathing in smoke may make your symptoms worse. Smoke can make you more susceptible to respiratory infections, like COVID-19. Some people most vulnerable to wildfire smoke, like those over 65 or with pre-existing conditions, are also those most at risk for serious impacts from COVID-19.

There were already limited ways to protect ourselves from wildfire smoke, and COVID-19 makes it even more challenging.

Here are steps you can take now to protect your health: 

• Stay informed about current and forecasted air quality here on the blog and your local clean air agency’s website.
• Reduce outdoor physical activity
• Stay indoors when it’s smoky and keep indoor air clean
• Close your windows and doors to reduce intake of smoke. However, ventilation is good for helping prevent COVID-19, so when air quality is good, open them to get fresh air and reduce potential viral load.
• Improve filtration of indoor air in your home and create a clean air room where you spend most of your time. Making your own box fan filter can be a less expensive option to filter air and improve indoor air quality in a single room.  Filtering indoor air is an effective way to reduce fine particles from wildfire smoke. It can also provide some protection from COVID-19, but this alone is not enough to protect you from COVID-19.
• Avoid burning candles or incense, smoking inside, frying or broiling, or vacuuming (unless your vacuum has a HEPA filter).
• Wear your cloth face covering to slow the spread of COVID-19. While cloth face coverings may help a small amount with smoke, they won’t filter out the fine particles or hazardous gasses.
• N95 respirators, if fitted and worn properly, can reduce exposure to wildfire smoke, but as the supply remains limited, these need to be reserved for workers that are required to wear them for their job.

For more information visit the WA DOH Smoke from Fires webpage.