Since the Fukushima nuclear disaster, you've probably heard me and other people talk about the radiation exposure we experience in everyday life. All humans, throughout history, have been exposed to background radiation produced constantly by the natural environment. Then there's added exposures from modern sources: X-rays and medical scans, living near power plants (both coal and nuclear, and the coal is actually worse), and flying in airplanes.

That last source of exposure works because the higher you get, the less you can rely upon Earth's atmosphere to shield you from radiation in space. It's the same reason why there's an increase in radiation exposure associated with climbing a mountain. All of these exposures are small. Small enough that most people don't need to worry about them. (For instance, a pregnant woman can safely take an airplane trip. You'd have to be a pregnant flight attendant, regularly working long-haul flights, before the exposures would start adding up to a quantifiable risk.)

But because we use these small-dose numbers to talk about relative risk and when radiation should and shouldn't scare us, it's interesting to know where they're coming from … and how accurate they are. That's why I was interested in something weird noticed by Ellen McManis. She operates a research nuclear reactor at Reed College in Portland, Oregon, and like many of us, she's curious about how much radiation people are actually being exposed to as a part of everyday life. Unlike us, however, McManis actually has access to things like dosimeters. With the help of her colleague, Reuven Lazarus, she recently took one on a cross-country plane flight—from Portland to DC, with a layover in Chicago. To her surprise, she found that the dose her dosimeter registered was actually a lot lower than the dose she'd been expecting.

I was using a RADOS RAD-60 dosimeter, which gives you an instant reading of how much radiation you've been exposed to while the dosimeter is on. We use them for visitors and people who don't have their own permanent dosimetry yet. Over the course of ~5 hours on the plane, I got a total of 0.3 millirem (or 3 microsieverts). I usually see a number quoted of 1 millirem per hour [for airplane exposure], or 3-to-5 millirem per cross-country flight, so that's an order of magnitude lower than expected.

Now that is interesting. If you look at Randall Munroe's Radiation Dose Chart (my favorite source for putting these small doses into context), you'll see that his well-researched numbers estimate an exposure of 40 microsieverts (the same thing as 4 millirem) for one cross-country plane flight. McManis' real-life reading was definitely a lot lower than the go-to estimate.

But why?

The truth is that McManis didn't really know. Her dosimeter was recently calibrated. She also checked it against a known source of radiation in the lab, and had turned up a result that was completely normal, so it seemed like this wasn't an issue of a wonky dosimeter.

Luckily, off-duty nuclear scientists aren't the only people taking measurements of in-flight radiation exposures. The official estimates, the ones used by people like Randall Munroe, come from an organization called the French Institute for Radiological Protection and Nuclear Safety.

Back in 1996, the European Union started counting radiation exposure on board airplanes as an occupational safety hazard. Remember, travelers generally don't have anything to worry about. But, for people who work on airplanes, the risk is large enough to be worth paying attention to, especially on certain routes. EU-based air crews are limited to 100 millisieverts of exposure every 5 years, and 50 millisieverts in any given year*.

How do they track that? You could, theoretically, give a personal dosimeter to every person working onboard an airplane. But that gets expensive, for reasons we'll talk about later. Instead, the EU has chosen to manage this with a system based on computer models—models that have been verified against more than 10,000 hours worth of real-world dosimeter readings.

It's called the Sievert System, and it works because the sources of radiation at 30,000 feet are fairly constant. Subatomic particles come from the Sun and from deep space to bombard our atmosphere. Reactions between those particles and our atmosphere produce secondary particles. Those secondary particles penetrate airplanes, and our skin, where they can damage our DNA.

There are factors that can alter the dose. Solar activity, for instance, means an increase in subatomic particles striking the atmosphere. Altitude matters, because the higher you are, the less atmosphere there is to protect you. Finally, latitude is also important. The particles penetrate our atmosphere more easily at the poles, says Jean-François Bottollier-Depois, head of the External Dosimetry Department at the French Institute for Radiological Protection and Nuclear Safety. By the time you get to 60 degrees latitude, he says, you will be experiencing a dose 2x as high as that at the equator. (Again, remembering that we're talking about very small doses.)

But these are all issues that can be factored into a computer model. All you need to know is the routes a pilot or crew member will fly in a given month, and the level of solar activity. The Sievert System uses that information to calculate monthly exposures for individuals.

Bottollier-Depois says the System also checks its work. Back in the early 90s, his team tracked the doses received by cosmonauts aboard Mir, so they know what the dose is in space. Earthside, they sent dosimeters on numerous flights, choosing a variety of routes, and taking measurements in different locations on the planes. They also used multiple dosimeters on each flight, so they could be sure that the dose recorded was accurate. And they still do these practical tests today, updating the Sievert System database to account for long-term changes in solar activity.

With all that experience under his belt, Bottollier-Depois had a pretty good idea of why Ellen McManis' measurements came out so wrong. In fact, it has to do with why the EU chose a model-based system, rather than real-time, individual dosimetry. All dosimeters are not created equal.

"If you use a classical dosimeter, it is measuring photons and electrons, but those account for less than 40% of the total dose aboard aircraft," he says. "The difference comes from the fact that you have other particles like neutrons, and those represent most of what you receive in a dose aboard an airplane. They can't be detected with classical dosimeter. You need very specific technology for that."

Expensive, specialized dosimeters pick up the particles that are most common at flight altitudes. Normal, old dosimeters don't. To McManis, that difference makes a lot of sense.

"I was using a personal alarm dosimeter that relies on ionizations to work, and neutrons don't ionize things," she says.

For more information, check out these links:

French Institute for Radiological Protection and Nuclear Safety — How far advanced is research on the health effects of low doses?

Sievert System page — You can calculate your own flight exposures here, and learn more about how the system works. (Heads up: In my experience, the site is often not working. If it won't load, check back in a couple days.)

*This story, as originally written, contained a typo. Pilots and airline crew are not limited to 100 microsieverts of exposure every 5 years, but 100 millisieverts. That's a big difference and it led to some confusion. My apologies. Thank you to Zac Labby for bringing this problem to my attention.

Image: airplane, a Creative Commons Attribution (2.0) image from freakland's photostream