This view of Pinatubo Volcano on the island of Luzon, was taken in July 1993, two years after it erupted. Scientists were able to correctly predict the eruption and save lives.

Image courtesy of Philippine Institute of Volcanology and Seismology.

Volcano forecasting depends on data like these:

In fact, Luhr argues, the problem with forecasting is not gaps in the science, but a lack of money to pay for equipment and people to gather data on volcanoes between eruptions. This "baseline" data helps them interpret the signs of changes when magma starts flowing at the start of an eruption.

Luhr points out that volcanologists have made several accurate forecasts in the past few years: "Pinatubo," he says, "was a great demonstration that if you get enough instruments and spend enough money, you can get the baseline information that can make a prediction possible."

Worth every cent
"I don't know how much was spent to evacuate [the area around Pinatubo] but it was a wonderful success," says Luhr. "There was a lot of money, and a lot of instruments, and scientists who were able to interpret the conflicting signals and make a judgment to make a call." Luhr adds that the 1980 eruption of Mt. St. Helens was also forecast correctly. But with so many volcanoes so near to population centers -- Mt. Rainier near Seattle, Mammoth Mountain in Southern California, and Popocatépetl near Mexico City all jump to mind -- the need for good forecasts is great and increasing.

Yet not all volcanoes receive the attention that Pinatubo did, which benefited from concern aroused by the proximity of two large U.S. military bases. "There are 1,500 active volcanoes around the world, and only a handful are adequately monitored," Luhr notes. Unfortunately, he adds, memories are short. After an eruption, governments may recognize the importance of spending money on volcano forecasting. "But if two years go by and there is not an eruption, it's off the political agenda."

Williams, who was badly burned in Colombia a few years ago studying an eruption in which six colleagues died, is trying to perfect and test a remote sensor for volcanic carbon dioxide emissions. He notes that volcanologists already read sulfur dioxide emissions, which tend to rise just before an eruption. But since carbon dioxide emissions should theoretically peak before sulfur dioxide, a remote sensor for carbon dioxide could provide safer and more timely warnings.

The gases are both taken up in magma as it forms deep below the crust, Williams says. But because carbon dioxide is far less soluble in magma than sulfur dioxide, as the magma rises and its internal pressure falls, carbon dioxide comes out of solution first, giving the warning signal. At least, that's the theory Williams is trying to test with his new remote sensor.

Our sensors says it's time to take a gander at some volcanic phenomena. Ready to visit the mondo bulldozer?