A US researcher is using hydrophones to identify and plot the courses of river traffic.
Alexander Sutin at Stevens Institute of Technology in New Jersey has developed algorithms that isolates the sound of individual boats, then tracks them based on time difference between several hydrophones.
"Classification parameters can be used like fingerprints to identify to identify what 
class a ship is," says Sutin who tested the system in the Hudson River.
The propellers of slow-moving boats like barges, for example he said, generate low-frequency modulation, while fast-moving speedboats produce high-frequency modulation.
The research was presented this week in two talks at a meeting of the Acoustical Society of America in Texas: Cross 
correlation of ship noise for water traffic monitoring and Passive acoustic classification of vessels in the Hudson River.