National Geographic News
Investigators inspect a badly burned building at 488 Quincy Avenue in Quincy. The cause of the fire is believed to be arson.

Investigators inspect a badly burned building in Massachusetts, suspected to be caused by arson.

Photograph by Aram Boghosian, The Boston Globe/Getty Images

Brian Clark Howard

National Geographic

Published September 7, 2013

The Rim Fire (see pictures) that recently ravaged Yosemite National Park was started by a hunter's illegal campfire, according to authorities.

To determine the cause of such a fire, investigators rely on a number of observations and techniques, from telltale signs of flames to satellite imagery and chemical tests.

To better understand the science, National Geographic spoke with Richard Meier, a fire investigator in Sarasota, Florida, with John A. Kennedy and Associates, which analyzes fires and explosions for insurance companies and trials. Meier is also a staff member at the National Association of Fire Investigators, and the organization's treasurer.

How does one become a fire investigator?

There's a lot of different ways. In my case I spent about 23 years in engineering, in the manufacturing industry, which is a good thing: I used to make things, so I know how they work. A lot of people start off as firefighters and then become investigators.

What kind of training is required?

I started training a few years before I came to Kennedy. I started attending seminars and webinars. The National Fire Protection Association maintains a list of subjects that investigators must know to be certified. These include fire chemistry, fire dynamics, documenting the scene, and so on.


How do you determine the cause of a fire?

Science had nothing to do with fire investigation up to the late 1970s and early '80s. Before that it was based on old wives' tales. But now we use the scientific method.

The first thing is to find out where the fire started, which is what we call the "origin." The "area of origin" can be a broad area, like a room in a house ... The "point of origin" is the smallest area you can identify, which can be as small as a thimble or as big as a room. In the case of a wildland fire, you're probably talking [about] an area of origin around a half acre to an acre, and a point of origin the size of a campsite or a campfire.

After you figure out the origin, you look for things that could have served as an ignition source: a campfire, a flare, hunting ammunition (rare but sometimes happens), or a curved bottle that's left in the sun and acts as a magnifying glass. You have to test each one, at least logically, and determine if it's a competent ignition source for the fuels that you have.

Fuels could be dried grass or pine needles, which take a lot of energy to get started. You can light a piece of paper with a match, but you can't light a two-by-four with a match.


An investigator looks through the Oakridge, California mobile home community following a wildfire fueled by strong winds in the Sylmar area of Los Angeles November 16, 2008.
An investigator looks through the Oakridge, California mobile home community following a wildfire fueled by strong winds.

Photograph by Phil McCarten, Reuters


How do you figure out the point of origin?

There are typical fire patterns, which you can see by looking at the way things get burnt, discolored, et cetera. For example, in a house you get V-shaped patterns on the wall that show the direction of the fire. You put those things together and start tracking back.

The amount of damage is an indicator, and witness statements are sometimes good. Sometimes you're lucky enough to get video. In one case I had video that was reflected off a piece of equipment, and it got us [to focus on] a small area in a large warehouse.

With wildland fires, aerial photography can be good. You can combine that with which way the wind is blowing. Satellites are being used more and more.

How is lightning ruled out as the cause of a fire?

The first thing you do is check the weather. Was there any lightning that day? Storms? Clouds? You use a process of elimination. If it was fair and sunny, you rule out lightning.

If you suspect a campfire, were there any witness accounts of people camping in the area? Anyone issued a camping permit? Anyone see a campfire before the forest fire?

Do you always look for "ignitable liquids," the industry term for flammable fuels like gasoline?

That depends. Generally I don't, but I mostly investigate accidental fires. Occasionally I investigate a fire that is accidental but someone is calling arson.

The National Fire Protection Association has a peer-reviewed document, NFPA 921, that serves as a guide for fire and explosion investigation, based on the scientific method. You look at the patterns first. If you see multiple points of origin, then you do chemical testing. [But] if you just start chemical testing, there are a lot of things that can give you a false positive, such as insecticides.

How are fires classified?

There are four official causes of fires: natural (such as lightning), accidental (such as when a log rolls out of a fireplace), incendiary (when the fire was set with intent for harm), and undetermined. It used to be that a lot of investigators didn't like to say "undetermined," but sometimes that's the right answer and sometimes we report that now.

How do you determine if a fire was set on purpose?

Law enforcement will look at behavioral factors, such as whether someone has way more insurance than what a building is worth, if they're behind on their mortgage, if they told people they were going to burn their house down, and so on. Yes, there are stupid arsonists out there.

[As I said,] I generally investigate fires that are accidental. But we're getting into civil litigation. We're looking for responsibilities, like a defective product—wiring or an electronic component that caught fire. We also see things used improperly, such as a space heater in a garage full of gasoline cans.


National Geographic also spoke with Michael Sigman, the director of the National Center for Forensic Science at the University of Central Florida in Orlando. The center conducts academic research on the causes of fires and supports forensic fire investigators around the country with information and a database of materials.

How do investigators determine the causes of a fire with chemical methods?

Investigators on the scene take samples of debris and send it to a lab. There, you have to heat the material in an oven. Any ignitable liquid residues left will go into a vapor phase. These are collected, commonly with carbon strips. The liquid is taken off the strip with a solvent, then analyzed in a gas chromatograph with a mass spectrometer attached.

Once analysis is done, [investigators] compare the results to standard protocols. Unlike on TV shows, their job is simply to determine if the debris [contains] ignitable liquid. If [it tests] positive that doesn't mean it was arson. [But the fire investigators] will try to assess what class of liquid it is, such as gasoline, petroleum distillates, oxygenated solvents, et cetera. And that will go to the authorities investigating the case.

What role does your center play in the process?

We don't do testing for specific cases, but we developed an online database that labs use when they're analyzing data from cases. We provide information on reference samples for different materials.

We also do research on topics important to the industry. For example, we've taken 50 ignitable liquids, evaporated them to various levels, and then reanalyzed that data, so investigators can start to understand how evaporated their sample might be.

If you have bacteria present, [that] can eat up some of the ignitable liquid, which can distort the results of a gas chromatograph. We're looking at how biological degradation affects data.

How do you determine if a fire was set on purpose?

That's [the investigator's] role, not our role, which is keeping the science pure. You could find gasoline in fire debris, but maybe someone was cleaning the carburetor from their car on their kitchen table, so the presence of something doesn't tell you why it was there.

Follow Brian Clark Howard on Twitter and Google+.



Popular Stories

The Future of Food

  • Why Food Matters

    Why Food Matters

    How do we feed nine billion people by 2050, and how do we do so sustainably?

  • Download: Free iPad App

    Download: Free iPad App

    We've made our magazine's best stories about the future of food available in a free iPad app.

See more food news, photos, and videos »