Navigating smoke

Given the many forest fires that have been burning out of control in the western U.S. I have received many inquiries about how to deal with smoke from an aviation perspective. First and foremost, smoke can present a hazard to all pilots, but especially those flying under visual flight rules (VFR).

Pyrocumulus near Los Angeles (photo provided by Chris O'brien)

Smoke lowers visibility, not only at the surface, but aloft as well. It is not usual for smoke to lower flight and surface visibility to below 1 statute mile which makes flying VFR dangerous. Even under instrument flight rules (IFR), visibility may be in the low IFR flight category and below published minimums for some airports. Even more concerning is flying at night and in mountainous terrain.

So what's a pilot to do? Of course, if you don't have to fly, that's likely the best option. Second, if you do decide to make the flight it's best to be on an instrument flight plan (VFR not recommended). Also, if you have oxygen on board, consider using it even below 10,000 feet. In fact, wearing an oxygen mask is always a good approach. All smoke contains some levels of carbon monoxide, carbon dioxide and other particulate matter. Hemoglobin bonds with carbon monoxide 200 times more readily than it bonds with oxygen and often produces hypemic hypoxia. Depending on what is actually burning at the surface, smoke can contain a variety of different chemicals, including aldehydes, acid gases, sulfur dioxide, nitrogen oxides, polycyclic aromatic hydrocarbons (PAHs), benzene, toluene, styrene, metals and dioxins. None of these are good to breathe especially if you have health issues (also consider your passenger's health).

The smoke starts off as eddies in the boundary layer. But then some of that air gets mixed further up above the boundary layer into the free atmosphere, and then it encounters stronger horizontal winds. Smoke can travel thousands of miles. In fact, some of the smoke from the western fires has even reached the east coast at this point, albeit, in low concentrations of particles.

In the early morning hours the atmosphere around the regions where the fires are burning is often fairly stable near the surface. That will trap some of the smoke keeping it close to the surface. The fires burn so hot that they often produce their own convective updrafts along with "clouds" called pyrocumulus, pyrocumulus congestus flammagenitus and cumulonimbus flammagenitus (with lightning) that carry the smoke high into the flight levels. These "clouds" don't produce any rain, but those updrafts can contain severe turbulence, that manifest as strong surface winds, which can exacerbate a large conflagration.

However, later in the day as the sun starts to heat the ground, that smokey air higher up moves downwind it gets mixed down from that turbulent mixing that takes place in the daytime. And that's how you get some of that smoke to show back up now near the surface at distances hundreds of miles from the origins of the fires.

Look for FU in the forecast or in the observations. FU is an abbreviation for smoke from the French word fumée as you can see below from this TAF from Lander (KLND).

What about using Model Output Statistics (MOS)? Here's the bad news. Unfortunately, MOS doesn't account for smoke in the visibility forecast. The terminal aerodrome forecast (TAF) is much more reliable in situations where smoke is expected in the terminal area. Those are issued by highly trained forecasters who can account for the effects of smoke. The Localized Aviation MOS Program (LAMP) does advect observational data such that in the first few forecast hours you will see the LAMP pick up on lowered visibility reported at airports, but beyond those few hours, it will quickly tend to discount the effects of smoke since it doesn't really have data to support this phenomenon. Certainly this is an area of research that will eventually integrate smoke into the MOS forecasts in some future release.

I like to use the smoke forecast from the High Resolution Rapid Refresh (HRRR) model which you can find here. HRRR-Smoke uses infrared (IR) satellite data to start. We know that fires create heat anomalies and that will show up nicely on IR satellite data. So it's not just about the smoke. Using this information means that the model is determining where the source of the fires are located. Once that information is known, it relies on changes in temperature, wind, water vapor, and precipitation to predict where the smoke will eventually end up in the atmosphere. Keep in mind that this is an experimental forecast.

The HRRR-Smoke is updated hourly and produces a forecast out to 18 hours for most hours, but for hours divisible by six (00Z, 06Z, 12Z & 1) it can provide a forecast with longer lead times. There are four different loops that can be used to include near-surface smoke, 1000 ft AGL smoke, 6000 ft AGL smoke and vertically integrated smoke. It's a good idea to look at all four.

For departing or approaching an airport the biggest concern is the conditions that might occur when landing or taking off. Near-surface smoke gives you smoke concentrations at about 8 meters (~26 feet) above the ground. As shown below, this is indicated on a pale-blue-to-deep-purple color scale at the bottom of the forecast map. As you might expect, California is currently covered in a smoky haze—purple and red is really bad, while light blue indicates fairly low concentrations (measured in micrograms per cubic meter of air). You can see that smoke has traveled clear across the country although these are very low concentrations.

Instead of measuring smoke around 8 meters off the ground, vertically integrated smoke is modeling what a 25-kilometer-high column of air looks like over any given location. The best way to think of this as the smoke that you can see covering the entire sky, versus the smoke near the surface that you can smell. As you can see below, smoke covers most of the country. The scale is a bit different in the magnitude of the numbers, but warmer purple-red colors are still very bad and the cooler pale blue colors represent much lower concentrations.

Even with these forecasts, on any given day it's extremely difficult to get a sense of what flying conditions will be like at cruise altitude. How high do you need to climb to be on top of the smoke assuming you can fly high enough to even get on top? There isn't a good answer. Certainly in regions where high concentrations exist using the vertically integrated smoke forecast, expect smoke from the surface well up into the flight levels. In other regions, smoke can often top out at 15,000 feet or so with higher concentrations below. It just depends on the current conditions including wind direction and atmospheric stability.

The best strategy is to look for pilot weather reports that often will point out where the top of the smoke is located. As with all pilot reports, when the smoke is really bad, pilots will often avoid flying through the area. If you do happen to fly on a smokey day, please take a few minutes to document what you experienced and submit a report.

Most pilots are weatherwise, but some are otherwise.

Scott Dennstaedt

Weather Systems Engineer

CFI & former NWS meteorologist


Our corporate headquarters is located in Charlotte, NC


  • Twitter Clean
  • YouTube Clean

© 2019 by AvWxWorkshops Inc