Volcano Watch — Seismographic networks and locating earthquakes

Release Date:

You hear a low rumbling sound; the walls of your house shake a little; objects on the shelves skip around, maybe even fall off the shelf. Was that an earthquake?

You hear a low rumbling sound; the walls of your house shake a little; objects on the shelves skip around, maybe even fall off the shelf. Was that an earthquake?

If you're on the Big Island, there is a very good chance that it actually was an earthquake. Each year at the Hawaiian Volcano Observatory, we receive reports of several dozens of earthquakes felt by residents of Hawai'i County. With our seismographic network, consisting of very sensitive motion-sensing devices, we detect many thousands of earthquakes during the same time period.

Among the basic information we use to describe an earthquake are its location and time of occurrence. This information can help us understand what might be happening to cause earthquakes within the volcanoes or beneath the Big Island. For large earthquakes, knowing the earthquake's location can also tell us where the most severe damage might have taken place and direct our emergency rescue and recovery efforts. It might also determine whether we should anticipate a tsunami or not.

Earthquakes occur at depths within the Earth that do not allow us to actually see them or the processes related to them. We must rely on observations made at ground level to determine what took place miles beneath the Earth's surface by developing rigorous models to explain our observations.

Fortunately, it has long been known that there are different types of waves generated by an earthquake that radiate away from the earthquake source, or hypocenter. These waves travel at different speeds, depending on the type of wave and the properties of the Earth where they pass through.

Using our seismographic networks, we record these waves as they arrive at our stations and make precise measurements of the wave arrival times. For the stations farther from the earthquake source, the waves arrive later than they do at stations located closer to the source. If we know how fast the waves pass through the Earth and the distances between the stations in our seismographic network, then it is possible to determine the distances between each of the stations and the earthquake that generated the waves.

So, locating an earthquake is quite straightforward in principal. Practically speaking, however, the job requires many steps and is complicated because the Earth is complex and, again, we are not able to observe directly the paths that the seismic waves follow.

Our seismographic stations continuously send data to HVO via radio. We gather the radio signals here, convert them to display them on our seismographs, and process them on our computers.

We run our data through a large number of automated processing steps. These include detecting possible earthquakes and estimating earthquake wave arrival times, sometimes hours before our data analysts are able to view the earthquake data. We continue to work on our automated procedures to improve the accuracy of our posted earthquake data.

The first step in locating an earthquake is to use a model that adequately describes how seismic waves pass through the Earth. We can use information from explosions for quarrying, as well as specifically for exploration, to independently determine seismic wave speeds. We are also able to use earthquake data to develop these models, and one of our current projects is to further promote the use of earthquake data to improve these Earth models.

The next step is to offer a tentative model for the earthquake. Most simply, this is done by specifying four numbers: earthquake latitude, longitude, depth and time of occurrence. To determine precise earthquake locations, we require data from greater than four stations, and we typically strive to use as much of our network as possible. Our precision is improved with widely distributed stations, as well.

From this initial location, and using our Earth model, we compute the times when the waves should arrive at our stations. Comparing these computed times with our measured arrival times, we make adjustments to the initial location and origin time to reduce the differences between the computed and measured arrival times. This process continues until our adjustments are small enough and the measured arrival times are reasonably fit by the model calculations. If we use different Earth models, our calculations will result in different earthquake locations.

With the advantages afforded by modern computers, these tasks are undertaken quite routinely. It is essential, however, that our experienced data analysts review the results of the automated procedures and add more arrival times and information to determine the earthquake magnitude. Our final verification of the earthquake report is not complete until as much of the data that we are able to use are incorporated into the model calculation.

Volcano Activity Update

Eruptive activity from the Pu`u `O`o vent on the east rift zone of Kīlauea Volcano continued unabated during the past week. About 50 percent of the lava output is flowing through the tube system and entering the ocean at Kamokuna. The other half is feeding a surface flow that is slowly approaching the coast 500 m (.3 mi) west of Kamokuna. The public is reminded that the ocean entry area is extremely hazardous, with explosions accompanying frequent collapses of the lava delta. The steam cloud is highly acidic and laced with glass particles.

Two earthquakes were reported felt since last week. A resident of Hawaiian Ocean View Estates in Ka'u and a resident of Aloha Estates in Puna reported feeling an earthquake at 12:53 p.m. on Friday, August 21. The magnitude 4.0 earthquake was located 15 km (9 mi) southeast of Honuapo at a depth of 33.4 km (20 mi). A magnitude 3.1 earthquake was felt in the Kīlauea summit region at 9:59 a.m. on Monday, August 24. The earthquake originated in Kīlauea caldera from a depth of 3.4 km (2 mi).