Moore and his group did their study at the request of the National Park Service, prompted in its turn by a committee of Native American tribal organizations, for whom the bridge is a sacred site. Last March they brought four seismometers, which measure vibration, to the bridge. A researcher rappelling down onto the span from a cliff placed two on the structure itself, and the others were deployed on either side of the bridge, some meters away. The researchers collected readings for 22 hours, and build a mathematical model of the structure using the measurements they gathered, and information about its shape.

Right away the data showed that the arch was picking up a range of vibrations. In the juddering lines of the seismograph readings they could watch the winds along the canyon dying down over the day, as well as the reverberations of the waves from Lake Powell. And “we could see things in our data straight away that looked like earthquakes,” says Moore. Two small local earthquakes rippled through during the measurement period, but there was one set of vibrations that looked different, with a much narrower range of energy that suggested it had traveled a long way. This appears to be from an earthquake induced by drilling to store waste water from oil production on the Oklahoma-Kansas border, Moore says. “The most interesting thing is just the recognition that this distant event is felt, is absorbed by Rainbow Bridge ... Even here in this incredible remote place, this earthquake from Oklahoma is rattling the bridge.”

The distant earthquake and the vibrations from Lake Powell, which is an artificial reservoir, activated the bridge at one of its resonant frequencies, the group notes. It's not clear what that means for its long-term stability. But it is a reminder that our activities are not anywhere near as constrained in space as we like to think they are. The model the team made provides a benchmark for the bridge going forward, however—future measurements will reveal whether the bridge's properties are changing as a result of vibrational damage. “We're poised to come back and re-measure the resonant properties,” says Moore. If a structure develops cracks and grows softer and less stiff, its resonant frequencies will drop. “If we come back in five years and all the resonant frequencies have dropped by, say, 10 percent, then we'd have some data to suggest that the arch has been damaged during that time,” he says. Moore hopes that this information, which his group is gathering about numerous natural bridges in the Southwest, will help the Park Service and other management organizations make decisions that will help extend the lifetimes of these magnificent structures.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.