Page 2 of 2
To be useful to first responders, models would have to take into account individual buildings, at the very least. In this scale, the smallest details of the city start to make a difference. In a city like New York, the long corridors of tall buildings create wind canyons, which determine how air moves. But two buildings of the same height create different turbulence patterns than two of staggered heights. Air also moves differently around tall buildings than it does around shorter ones, which it can also go over. It can move more quickly up along the sides of sunny buildings, where heat helps it rise. The vast, underground subway network and its ventilation system complicates things, but so do fountains, which can affect the air’s buoyancy. The two-foot cornice of a building can change the pattern of the wind. So can trees, roads, courtyards, gardens; the impact can extend to within a meter, or within a hundred meters.
To make a model able to reckon with that level of detail requires extraordinary computer power, which has only recently been brought to bear on the urban weather studies. (For many years, the best models that scientists studying air and wind movement had were built in wind tunnels: plexiglass areas, as small as an East Village apartment or as large as a high school gym, where researchers would build a physical model to scale and, essentially, turn on a fan and make measurement. )
“Because we have the computer power, we can do the same kind of things, but we can run them over and over and over again,” said Teddy Holt, a Naval Research Lab* scientist who has worked on modeling. “It's the computer power that has allowed us to such high resolution that we can start predicting these kinds of things.”
By 2005, that initial meeting of the Urban Atmosphere Observatory had led to two experiments, one in the Madison Square Garden area, another in Midtown, in which scores of scientists and their assistants released and then monitored the path of an odorless, colorless gas. The media showed up but quickly lost interest, said Steve Hanna, one scientist involved in the experiment.
“They apparently were expecting visible smoke and spinning instruments,” he wrote in an email. “But the tracer gas was invisible and was released in tiny quantities from a small tank and hose which just sat there and did nothing. Similarly the winds were observed by sonic anemometers which also do not spin around and just sit there, not moving. The gas samplers were small stationary hoses and boxes which we carried around and set up on light posts and rooftops in what looked like beer-coolers. It was not like a Weather Channel special showing meteorologists chasing tornados and houses being blown apart.”
Two years later, when the project finished up, its leaders made no effort to trumpet its accomplishments. The City College of New York inherited the network of equipment that had been procured for the project. The government decided to keep portions of the data collected close to its chest, particularly those related to the subway. One of the main results of the tests was to show how limited the current models for predicting air flow were. As Teddy Holt, who was involved with the project put it: “Oftentimes it went places they thought it would go, but it also went places they thought it wouldn't go.”
The federal government is still working on the product it promised to the city’s emergency management office. The work begins with the models atmospheric scientists have created of the city’s air system. They are like maps of the city’s atmosphere: if you plug in certain starting conditions—location plus wind direction—they’ll show you where a released gas will go next. These models will eventually feed an application that emergency responders could load on a handheld device and, after plugging in those few pieces of data, get a read on where a toxic release is bound for. But creating the applications is arduous: the most detailed models run slowly, over a few hours on supercomputers, so to create a quick-response program, scientists have to run, in advance, a large range of scenarios, creating a library of look-up tables that will be more quickly available. And since New York is so complicated, the government scientists working on the project have essentially put the project here on hold, by focusing on creating the application for D.C.-based responders first.
Even these most sophisticated models do not account for things like the subtle effects of the heat coming from the sunny side of a building, for instance. But atmospheric scientists around the world are still very much interested in understanding and modeling what happens in the street canyons of cities like New York. They want to make pedestrians more comfortable, by creating systems that help whisk away heat and pollution from the city streets. They want to make cities more sustainable by understanding how they interact with sunlight, and trap heat, in order to fix those tendencies and abate climate change.
The funding for scientists pursuing this sort of work has shifted away from homeland security and back to environmental concerns. Steve Hanna has found that at conferences, everyone seems to be “becoming an urban climatologist.” Stuart Gaffin, the Earth Institute scientist, is studying white roofs.
“The time is right to revisit Lower Manhattan as another experiment site, because it's so different from Midtown,” Pullen says. “These types of things tend to go in cycles.”
*corrected from the original
