May 31, 2007
A weekly feature provided by scientists at the Hawaiian Volcano Observatory.
Recent Mauna Loa "prediction" does not change HVO assessment
A few weeks ago, each of the local newspapers carried a story about a study, released in the journal Science, predicting that the next Mauna Loa eruption would occur in its southwest rift zone. The stories were titled in various interest-drawing ways, like "Mauna Loa's next eruption toward Ka`u" and "Mauna Loa under pressure: Mauna Loa's next eruption likely along Southwest Rift, but when is less certain than where."
The study's main author, Dr. Falk Amelung, is quoted in the Honolulu Advertiser as saying, "If an eruption happens, it's most likely to happen in the Southwest Rift Zone." "Why didn't the Hawaiian Volcano Observatory pick up on this important development and warn the public before now?" you might ask.
The answer to that question lies in the definition of the word "important." Dr. Amelung's study is an important scientific contribution, but it is not important for lava flow hazard prediction in the same way that HVO issues warnings when an eruption is imminent. Dr. Amelung's prediction is based on a mathematical model of how Mauna Loa responds to earthquakes and eruptions. If the model is a good one, it could be used to predict the areas in which earthquakes and/or volcanic eruptions are most likely to occur. If the model fails, it probably failed because it was too simple. An analysis of its failures can be used to improve subsequent models.
The value of Amelung's model has yet to be assessed. It explains several things about Mauna Loa's past behavior, but several assumptions have been made. The real world is very complex, and simplifications are required for any model to be workable.
A good analogy is the multiple predictions about the correct answers to questions raised in the Season 3 finale of the TV production "Lost," also based in Hawai`i. Based only on data (scenes) in the finale, several possible answers exist for the question "Who is in the coffin?" Several viewers were so eager as to freeze-frame through the finale to try to enhance images of newspaper articles and otherwise compile evidence. They developed models based on what they had learned from the finale as well as their close observation of all the season's episodes.
The correctness of anyone's model will rest solely on whether it correctly predicted the coffin's occupant. Those of you who do not watch "Lost" could translate this to all the theories and/or predictions put forth about "Who shot J.R.?" a few decades back or who will win the 2007 World Series, the next Super Bowl, or the next "American Idol."
While HVO staff is very interested in Amelung's model, they won't be using its predictive capabilities until it's been proven correct a few times. Instead, we stick with the model that has worked for us in the past - history. Mauna Loa will behave in the future as it has in the past. Since 1843, nearly all of Mauna Loa's eruptions have started at its summit. About half of them have stayed in the summit, and the other half have moved about equally into either rift zone. Therefore, the chance of the next eruption of Mauna Loa moving into the southwest rift zone is about 25 percent. Amelung's model predicts a 100 percent probability, and Mauna Loa's history says he has a 1 in 4 chance of being correct.
One thing is for certain - Mauna Loa will erupt again. For the last few years, HVO has been reporting that Mauna Loa has been inflating at slow rates; HVO continues to watch closely. Studies like Amelung's may prove very useful once tested and verified. We all continue to gather clues, analyze patterns, and propose predictive models.
We may know the answer to the "Lost" questions before we know whether Dr. Amelung's model was right.
This past week, activity levels at the summit of Kilauea Volcano have remained elevated above background levels. The earthquakes were concentrated in the upper southwest and east rift zones. The summit caldera has been expanding, indicating inflation, since the beginning of 2007.
Eruptive activity at Pu`u `O`o continues. On clear nights, glow is visible from vents within the eastern half of the crater. Lava continues to flow through the upper portion of the PKK lava tube to where the Campout tube branches off about 1 km south of Pu`u `O`o. The original PKK lava tube below this point has been abandoned. A breakout from the upper, active section of the PKK tube has been sending lava downslope onto the upper flow field for the last few weeks. Glow at night from this area, seen from the coast, suggests that the breakout continues to be active.
The Campout tube continues to carry lava downslope, where it is feeding breakouts scattered across the coastal plain from near the base of the Royal Gardens subdivision to the ocean entry at Poupou. The Poupou entry has begun to coalesce into two separate entry points, each building a small delta, and is located within Hawai`i Volcanoes National Park. The Kamokuna entry, which was fed by the main branch of the Campout tube, is no longer active.
Access to the sea cliff at the ocean entries is closed, due to significant hazards. The surrounding area, however, is open. If you visit the eruption site, check with the rangers for current updates, and remember to carry lots of water when venturing out onto the flow field.
Four earthquakes beneath Hawai`i Island were reported felt within the past week. A magnitude-3.9 earthquake occurred at 10:51 a.m. H.s.t. on Thursday, May 24, and was located 6 km (4 miles) southeast of Kilauea summit at a depth of 1 km (1 mile). A magnitude-3.1 earthquake occurred at 11:05 p.m. on Friday, May 25, and was located in the same area at a depth of 3 km (2 miles). A magnitude-2.4 earthquake occurred at 9:50 p.m. on Tuesday, May 29, and was located 4 km (2 miles) southwest of `O`okala at a depth of 11 km (7 miles). A magnitude-2.2 earthquake occurred at 6:07 p.m. on Wednesday, May 30, and was located 2 km (1 miles) northeast of Pu`ulena Crater at a depth of 2 km (1 mile).
Mauna Loa is not erupting. No earthquakes were located beneath the summit. Extension of distances between locations spanning the summit, indicating inflation, continues at steady, slow rates.
Updated: June 8, 2007 (pnf)