A Lucky Mistake

A mathematical error proved fruitful for Norden Huang-and other scientists trying to interpret data.

Norden E. Huang considers himself a lucky man. In 1995, he wrote a paper for Advances in Applied Mechanics in which he used-incorrectly, as it turned out-a set of mathematical formulations called the Hilbert Transform, named for David Hilbert, one of the 20th century's most important mathematicians. Huang had already submitted his paper when a colleague pointed out his error. He called the editors and asked for his paper back, but they found his conclusions intriguing, and instead gave him a two-month extension for revision.

"In trying to correct my mistakes, I stumbled upon this matter," says Huang, who retired in September as chief scientist for oceanography at NASA's Goddard Space Flight Center to teach at Taiwan's National Central University. By "this matter," he means a whole new way to analyze data-a set of algorithms he developed and later named the Hilbert-Huang Transform. When Huang presented his findings, a Johns Hopkins University professor who had been his graduate school adviser many years earlier pointed out its significance. "He said, 'This thing you happened upon happens once every 100 years,' " says Huang, his voice still full of wonder.

Of course, it wasn't just luck. Huang has since been recognized by a host of awards, and HHT is described as "one of the most important discoveries in the field of applied mathematics in NASA history." Huang-who decided to study fluid mechanics at Johns Hopkins be-cause the cargo ship on which he traveled from Taiwan broke down, and by the time he arrived on campus the research positions in solid and structural mechanics all were taken-originally used his method to analyze ocean waves. It has since proved useful at NASA in looking for planets and black holes, testing insulation tiles for the Space Shuttle Return to Flight project, and designing the next generation of aircraft.

Huang also helps scientists from universities and other government agencies-including, recently, researchers from Harvard Medical School-to answer previously impenetrable questions by using HHT. The method has been applied to climate change, bridge safety, submarine design, speech patterns, blood pressure, heartbeats and many other subjects. "I'm no expert of any field, I just help them interpret the data," says Huang with characteristic modesty.

So what-exactly-does HHT do?

The computer-simulated method is a tool for analyzing data from nonlinear and nonstationary processes with much greater precision and flexibility than was previously possible. "Once you write down a [mathematical] formula, everything is concrete," Huang explains. "It is no longer changeable." But few phenomena-natural or man-made-fit neatly into those constraints. When you measure something that changes at varying rates or is not uniform across space and time, things become much more complex. "HHT is trying to get over this kind of limitation to get meaning out of the data," Huang says.

Pondering how Earth's orbit determines climate cycles, what the placement of black holes tells us about the creation of the universe and how heartbeats resemble ocean waves-these diverse questions keep life interesting for Huang.