December 2, 1999
A weekly feature provided by scientists at the Hawaiian Volcano Observatory.
Does heavy rain influence eruptions? Maybe
The heavy rain that East Hawai`i experienced this week, particularly on Wednesday night and Thursday morning, suggested the topic for this week's column. Does heavy rain influence eruptions? If you get three volcanologists around a table and ask that question, one will say "yes," one "no," and the other "maybe." Then, after some discussion, they will all probably say "maybe." In other words, the topic is controversial and unsettled.
There is no doubt that rainfall has many indirect effects. Rain creates lakes, and eruptions through lakes are commonly explosive because of the mixing of magma with water. Groundwater, the result of rainfall infiltration into rocks and soil, may interact with magma in the same way as lake water. The details of these interactions are complex, and conditions have to be just right for explosions to be triggered. The 1924 explosions from Halema`uma`u, and many of those between about 1500 and 1790, result from interactions of magma or lava with groundwater or lake water.
Once a lava flow is underway, the presence of rain-fed lakes or streams can strongly influence the final appearance and even location of the flow. Water cools lava much faster than air, so that crust thickens rapidly. Barriers made of quenched lava can form within a flow, causing the flow to thicken or even be diverted. Or, the lava can granulate upon contact with water to form black glassy sand, so common along beaches on Kilauea and Mauna Loa.
Eruptions in wet environments can result in mudflows, which rush down mountain valleys and present some of the most devastating kinds of volcano hazards. For example, more than 22,000 Colombians were killed by a mudflow during the otherwise small eruption of Nevado del Ruiz on November 13, 1985.
These are some of the indirect effects of rainfall. But what about the direct effects? Here is where the uncertainty comes in.
The best documented study of the direct effect was made by Larry Mastin, a colleague at the Cascades Volcano Observatory, who studied the relation between storm rains and explosions or explosion-like seismic signals from the dome of Mount St. Helens in 1989-1991. He found that most of the 28 volcanic events followed within hours to, at most, days after storms, which dropped large amounts of rain or snow on the hot dome. Mastin interpreted this observation to indicate that the rapid precipitation caused the surface of the dome to cool rapidly and crack. Volcanic gas, constantly coming from great depth but trapped within the dome, was able to escape violently through the new cracks, causing explosions or explosion-like seismic signals.
Norm Banks, former HVO volcanologist, and his colleagues concluded that seismicity and ground deformation of Nevado del Ruiz volcano (Colombia) are correlated with weekly rainfall. Other workers have made correlations between rainfall and volcanic and seismic phenomena at volcanoes in El Salvador, Guatemala, Mexico, and Iceland. These are general statistical correlations, not specific to single eruptions.
Former HVO seismologist, Fred Klein, found a marginally significant statistical correlation between the start of eruptions and high rainfall and a somewhat better correlation between eruptions and seasonal precipitation. These correlations are not strong and again don't hold for specific events.
To be objective, it seems best to conclude that the jury is out on whether rainfall directly influences specific eruptions. In other words, we offer a definite "maybe." For what it is worth, we noted no effect on the ongoing Kilauea eruption by the heavy rain this week, even though our gage at Pu`u `O`o recorded 10 cm (4 inches) on Wednesday night and 15 cm (6 inches) for the 24-hour period ending Thursday morning.
Eruptive activity of Kilauea Volcano continued unabated during the past week. Lava is erupting from Pu`u `O`o and flowing through a tube to the southeast in the direction of the seacoast. Breakouts from the tube feed a series of perched lava ponds above Pulama pali. A broad `a`a flow originates from the lowest perched pond near the 650 m (2150 ft) elevation. As of December 2, the distal end of the flow was below the base of the pali at an elevation of 90 m (300 ft) and was 2-2.3 km (1.2-1.4 mi) from the coastline. A pahoehoe flow from a higher perched pond is located 900 m (2950 ft) east of the `a`a flow, and is visible near the top of Pulama pali at an elevation of 400 m (1300 ft).
On Wednesday, December 1, a resident of Pahala felt an earthquake at 11:57 in the morning. The magnitude-2.8 earthquake was located directly beneath Pahala at a depth of 4.5 km (2.7 mi).
The URL of this page is http://hvo.wr.usgs.gov/volcanowatch/archive/1999/99_12_02.html
Updated: 5 Jan 2000