By Joyce Kryszak
Amherst, NY – UB researchers are applying virtual reality and supercomputing to warn people about the potential danger from active volacanoes. The work is being funded under a three-year, $1.9 million grant from the National Science Foundation.
Scientists have for years been able to predict, with amazing accuracy, when a volcanic eruption will occur. But for the first time, technology is being harnessed to tell people how to get away from the wrath of that eruption -- the deadly flow of molten lava.
The multi-disciplinary technology combines data, simulations and communications to more accurately predict the path of lava flows. Public officials in risky volcano areas are linked, on line, immediately to 3-D images that outline the most likely paths of destruction. And UB geologist Michael Sheridan says that means knowing who should be evacuated and to where.
"What this will do will be to provide a number of scenarios with different levels of probability, so that people will be able to gadge is it going to be a small event, a medium sized event, or a large event -- and where will it happen," Sheridan said. "Not only that, but they will be able to see it come in and by clicking into their own location where they'll get a view of that location, featuring an animation in three dimentions of what could happen."
Initially, the UB team of geologists, engineers and mathmaticians is focusing on three Mexican volanoes, including its most active volcano in Colima. UB researcher Abani Patra, principal investigator on the grant, says UB is ahead of the pack in the race for geological forcasting because the program is unique.
"You will not get this collection of resources, expertise and people and everything togther, anywhere else," Patra said. "And what we have, very uniquely here, is collaborations with the actual people in the field who need this stuff."
The UB team says it will continue working to expand upon the success of the simulation program which can now predict the actual path of lava flow with about 50 percent accuracy.