Sniffing Out Threats

New York's mystery odor shows that modern sensors are limited in their ability to detect airborne chemicals.

On Jan. 16, the city of New York gave up.

It had been a week since an unknown, pervasive odor had seeped into parts of Manhattan. City officials decided to call an end to the search for the source of the smell, without having figured out the mystery.

By this point, the absence of any deaths or serious injuries assured the populace that the odor was benign, making it comedic fodder for late-night talk show hosts and Manhattanites who dislike New Jersey. But what if it had come from an intentional, nefarious release? With the chemical weapon detection technology currently in use and under development, how is it possible not to know the cause of an odor pervasive enough to close down several buildings and force the evacuation of a commuter train? The answer is that detecting airborne toxins and tracing certain smells are two different things, and both can be difficult.

"There's not a machine that can detect odors," says Ian Michaels, spokesman for the New York City Department of Environmental Protection. "There are machines that can detect substances in the air, but not odor. And there is a difference."

In urban centers across the country, the Homeland Security Department operates BioWatch, a nascent air surveillance system meant to provide early warning of a biological release. The system, criticized in some circles for taking too long to provide results, is now up and running in 30 cities, with a second-generation version in use in the 10 cities deemed to be at greatest risk of terrorism. But most biological releases-bacteria, viruses or other pathogens-would be unlikely to leave a smell. Some chemical agents, such as sarin, leave only a faint odor; others, including chlorine gas and hydrogen sulfide, come with a distinct smell. Chemical warfare agents are much more deadly indoors or in enclosed spaces.

Cities such as New York own and operate a slew of chemical sensors, the details of which local officials are loath to discuss for security reasons. But the problem the technology presents is that for the most part detectors test for the presence or absence of a specific set of chemicals or compounds. Testing for an unknown compound or precisely identifying all the compounds in the air is a completely different task.

"The air we breathe is not pure and lily white; it's full of things that can cloud our ability to [detect toxins]," says Dennis Perrotta, associate professor of epidemiology and biosecurity at the University of Texas School of Public Health in Houston. "On a day-to-day basis in a normal day of operations in the city of New York, there are plenty of other things that cause odors as well."

That's where the city's Department of Environmental Protection comes in. One of its many responsibilities is enforcing the city's rather strict odor control laws-the work of about 50 inspectors. The city's hazardous material teams have mobile detection units that scan for volatile organic compounds and other harmful substances. But if the number of complaints suggests a larger event, then the department sends some of its trained smell inspectors out to trace it. They take air samples but, Michaels says, they mostly use their noses and maps generated from resident complaints.

"You map where an odor may be moving; obviously, you can't see it," he says. "So, if you know at 8 a.m. it was in Staten Island, and at 9 a.m. lower Manhattan, 10 a.m. midtown, 11 a.m. upper Manhattan, then you have a pretty good idea of where it's moving, and many times where it came from."

The inspectors then can check manufacturing plants and other nearby facilities that could have been the source of a release. Ultimately, odor detection is as much about finding the source as analyzing the substances in the air, because of the complexity of the latter.

"You could smell an odor and you could say, 'it smells like burning rubber,' but that burning rubber smell could be caused by hundreds of different compounds in the air," Michaels says.

If a chemical release is not continuous, but limited to a brief, single moment-as could be the case in a chemical attack-finding the source would be much more difficult. By the time the release was noticed, the wind could have taken it in any number of directions.

Michaels says that for widespread cases, the department reaches out to federal agencies, such as the Environmental Protection Agency, and neighboring ones, such as the New Jersey Department of Environmental Protection. Officials with the New Jersey department say they checked 157 facilities and confirmed by Jan. 16 that none had been the source of the mysterious smell that befell Manhattan on Jan. 8.

Some theories suggest New York's mystery odor was mercaptans-a smelly substance added to natural gas so homeowners can notice a gas leak. Three Columbia University scientists wrote in a New York Times editorial that the smell was sulfur gases from saltwater marshes in the metropolitan area. Though the ambiguity can be unsettling, Perrotta says it is unreasonable to expect that inspectors and detectors could identify and trace every odor. "I believe there's not a real way to protect against all [potential airborne releases]," he says. "It's actually quite complicated when it seems like it should be simple."