By Steve Hamm
Mark Twain’s often repeated quip, “Everybody talks about the weather but nobody does anything about it,” captures perfectly humanity’s frustration with the vagaries of weather. Take heart, though. While IBM Researcher Lloyd Treinish doesn’t claim he can change the weather, he and his environmental science team at IBM Research promise they can do the next-best thing: help people make better decisions based on pin-point accurate weather predictions.
The team’s sophisticated weather prediction technology, nicknamed Deep Thunder, can help city managers prepare for severe storms; farmers plan the planting, irrigation and harvesting of crops; and electrical utilities respond to hurricanes or integrate alternative energy sources such as wind and solar into their grids. “When we think of the impact of weather on business, it’s enormous,” Lloyd says. “In this country alone, it’s estimated that a significant fraction of the GDP is dependent on weather conditions.”
Eventually, the team’s technology could be useful for individuals as well. Already, Treinish gets calls from IBM research colleagues who are planning backyard barbeques and want to avoid rainouts.
Deep Thunder is an example of a new generation of cutting-edge technologies that penetrate complexity by capturing and making sense of large quantities of data
If you want to learn more about the future of technology, download a free chapter of the book Smart Machines: IBM’s Watson and the Era of Cognitive Computing, by IBM Research Director John E. Kelly III, at cup.columbia.edu/static/cognitive
Lloyd didn’t start out wanting to be a weather scientist. Like many other Americans growing up in the 1960s, he dreamed of being involved in the space program. As a young boy in suburban Maryland, he frequently borrowed his father’s WWII binoculars for stargazing in the backyard before he got his first telescope. He recalls being in summer camp in the summer of 1969 and watching the first moonwalk with other campers crowded around a small black and white TV. “I was excited by what we didn’t know—the challenge of the unknown,” he says.
As Lloyd made his way through school, his keen interest in astronomy morphed into fascinations with related topics—planetary science and physics. He got bachelors and masters degrees from Massachusetts Institute of Technology, and planned on pursuing a PhD. at the University of Arizona, but a job opportunity that he could not refuse sent him off in another direction. He went to NASA’s Goddard Space Flight Center in Greenbelt, Maryland in 1978, where he developed advanced scientific data systems in earth and space sciences for the next 12 years.
Key elements of Lloyd’s work at Goddard were high-performance computing and data visualization—which caught the eye of IBM researchers, who were interested in using powerful computers to model and simulate large-scale natural phenomena. IBM recruited Lloyd in 1990 to pursue his work at the lab in Yorktown Heights.
The Deep Thunder project emerged out of collaboration with the US National Oceanic and Atmospheric Administration (NOAA), the parent of the National Weather Service. NOAA wanted to be able to predict the weather in particular areas with a high degree of accuracy. Their focus was the Olympic Games in Atlanta in 1996. Would they be able to predict weather conditions for particular events far enough in advance for officials to react and make adjustments? It became clear to the scientists that to be able to predict precise weather conditions in a particular place at a particular time, they’d have to produce advances in data acquisition, computing and modeling.
In a sense, this was the first of IBM’s Smarter Planet engagements—though it predated the strategy and moniker by more than a decade. It involved gathering a tremendous amount of weather data from sensors via a computer network, integrating the data, using math and a supercomputer to analyze it, and employing visualization technology to make the information useful on a Web site. The system correctly predicted thunderstorms that interfered with a sailing race and assured officials that storms forecast for the area would not interfere with the closing ceremonies.
Since then, Lloyd’s team has engaged with NOAA and other government agencies on additional projects, and with a wide variety of partners and clients, including several electrical utilities. The meteorological team now includes three people in Yorktown Heights, three in the China lab, one in the India lab and one in the Brazil lab—plus help from other scientists with expertise in computer science, hydrology and statistics.
Their highest-profile engagement so far has been with the city of Rio de Janeiro, in Brazil. After an unusually strong and unexpected storm resulted in severe flooding and loss of life in Rio in 2010, Lloyd’s team was invited in to develop weather modeling and simulation technology to help the city anticipate storms and prepare responses. First they created a meteorological model with one-kilometer resolution for the entire metropolitan area. Then they combined that model with a hydrological model designed to capture what happens when the water hits the ground and flows down the city’s steep hills. The results of their analysis are available to city managers via a specialized Web portal using innovative 2- and 3-dimensional visualizations. Today, using the system, city officials can see forecasts of local weather events up to 40 hours ahead.
Most recently, they expanded their capabilities by applying the results of the flood model to evaluate its impact on traffic planning. The goal is for operations managers to be able to identify needed changes in traffic flow to minimize the impact of a severe storm even before it starts to rain.
What’s next for Lloyd and the team? In truth, Lloyd feels like they’ve just scratched the surface of what’s possible. He sees many uses for their technology in public safety, energy generation and management, agriculture and even casualty insurance. “There are very challenging problems that nobody has tackled yet,” he says, but, he adds: “I see so much potential in how this technology can help people.”