A new detailed look at several years of swelling on Mauna Loa's Southwest Rift Zone suggests that the volcano's next eruption will be along the rift, which extends from the summit toward Na'alehu in the Ka'u District.

A study yesterday in the journal Science states that measurements from a Canadian satellite are providing telling clues about the volcano's next outburst.

"If an eruption occurs, we have a clear idea where it will occur. If an eruption happens, it's most likely to happen in the Southwest Rift Zone," said Falk Amelung, one of the study's authors and a geophysics professor at the University of Miami.

But Amelung, a former University of Hawai'i researcher, warned that predicting volcanoes is a tricky business. An earthquake could change the structure of the rock and provide a new route to the surface for molten rock.

There may be existing structural components that could prevent magma from pushing directly upward from where it is accumulating. The movement of magma from deep in the Earth's mantle to the volcano's magma storage areas could slow, reducing pressure and the likelihood of an eruption.

The last eruption along the rift, in 1950, sent lava flows down both the south and west sides of that long ridge. Three flows reached the ocean between the South Kona village of Pahoehoe and Miloli'i. Numerous homes were destroyed, along with a gas station and post office.

The eruption started at the summit and had repeated outbreaks lower on the volcanic ridge at the Southwest rift, covering 43 square miles with fresh lava.

A 1984 eruption did the same thing on the Northeast Rift Zone, launching flows that were headed for Hilo before the eruption ceased and the flows stopped well short of the city, with 18 square miles of land covered by new rock.

Mauna Loa, which has had seven flank eruptions in the past century, last pumped molten rock to the surface in 1984. Statistically, an eruption is overdue.

An eruption solely along the rift would be atypical, however.

"Historically, Mauna Loa eruptions start at the summit," said Jim Kauahikaua, scientist in charge at the Hawaiian Volcano Observatory. However, such eruptions often quickly migrate down the mountainside along one of the volcano's rifts.

HIGH-TECH BLACK HOLE

Kauahikaua is less willing than Amelung to predict the site of the next eruption, if only because scientists' high-tech tools — such as satellite measurements — are so new that there's nothing to which they can compare the new data. For instance, he said, no one knows whether the kind of swelling Amelung sees on the Southwest Rift Zone happened before the 1950 eruption, or was present before the 1984 eruption on the Northeast Rift Zone.

"We just haven't seen enough of these things," Kauahikaua said.

Amelung from 2002 to 2005 was able to precisely measure a region of swelling growing on the surface of Mauna Loa, about 3 miles southwest of the summit caldera. There is a bump on either side of the rift. The overall inflated area measures about 9 miles across. The ground has risen about 7 inches.

Most active volcanoes are assumed to have a magma chamber below the summit, and volcano scientists often calculate the inflation at the summit. But this inflation was off to the side — suggesting liquid rock was collecting in a new place, away from under the summit.

Because of the high quality of the satellite data, "we could infer very precisely where the magma was accumulating," Amelung said.

The team's theory about how this came about goes like this: A 1983 earthquake shook up the mountain, creating cracks and "unclamping" a geological region that had been under great stress. The unclamping — the term the geologists use — allows liquid rock to readily move into the region, while that would have been difficult when the rocks were under more stress.

Amelung said the 1984 eruption filled with rock, and essentially reclamped, the Northeast Rift Zone. So when magma began inflating the volcano, it moved into the only unclamped place left — the Southwest Rift Zone.

NO SIGNS OF URGENCY

The team's theory of the inside of the volcano is that both the semipermanent magma chamber under the summit and the magma storage area in the Southwest Rift Zone have independent connections to the mantle, but that they are also connected to each other in some way, so magma can move between them. They feel that the zone's liquid rock is contained in a complex of underground cracks.

But it doesn't mean a Mauna Loa eruption is imminent. Kauahikaua said the swelling of the volcano has slowed somewhat in recent months, and there is no significant change in underground tremor to indicate an eruption is in the immediate future.

One thing that could preface an eruption, Amelung said, is a major quake in the Southwest Rift Zone, associated with slippage along the fault zone. That could provide an avenue for magma to find its way to the surface in an eruption.

When or whether that might happen, no one is predicting, but both Amelung and Kauahikaua said the technology of monitoring volcanoes is getting better fast, and can't help but improve forecasting.

"If we continue to get these very precise measurements, eventually we will be able to predict with some more accuracy what will happen," Amelung said.

Amelung worked on the study with Stanford geophysicists Sang-Ho Yun and Paul Segall and Miami colleagues Thomas Walter and San-Wan Kim.

Reach Jan TenBruggencate at jant@honoluluadvertiser.com.