Categorized | Featured, Sci-Tech, Videos, Volcano

Volcano Watch: Super models help with volcano monitoring

A model of the magma reservoir 1.5 km (1 mi) below Halema`uma`u Crater at the summit of Kilauea Volcano. Image courtesy of USGS/HVO

A model of the magma reservoir 1.5 km (1 mi) below Halema`uma`u Crater at the summit of Kilauea Volcano. Image courtesy of USGS/HVO

(Volcano Watch is a weekly article written by scientists at the U.S. Geological Survey’s Hawaiian Volcano Observatory.)

No, Tyra Banks hasn’t been helping HVO install monitoring equipment at Kilauea. We’re talking about using computer models to investigate what is happening inside the volcano.

Scientists use many techniques to infer the processes occurring beneath active volcanoes, where we can’t directly see what is happening. At HVO, we use earthquakes, ground deformation, gas emissions, and geologic observations to understand what’s going on beneath our feet.

Using this monitoring information, scientists develop “models” to explain what is happening within the volcano. Even simple cartoons are models—simplifications that we use to better understand more complex systems. Basically, models are ideas or representations that we use to approximate nature. As additional data are collected, models can more accurately represent the real world.

Scientists at HVO are starting to use super computers to model interactions between different parts of a volcano. These new “finite element” models split a schematic volcano into thousands of small pieces, or “elements.” Each of these elements can be told how to act under certain conditions. For example, if it is heated, the element knows what temperature to melt at, and if it is squished, how to change shape. Each element also interacts with its neighbors, pushing or heating its surroundings. In this way, scientists can simulate an event—like the deflation and inflation (DI events) at Kilauea’s summit—to investigate how the rocks are heated and deformed as magma pushes towards the surface.

The models are built using information from HVO’s monitoring efforts, such as erupted volumes, earthquake locations, and surface deformation, so that they are as accurate a depiction of real events as possible. We can then use these models to investigate factors that we don’t know, like the volume and pressure of magma within the summit magma reservoir or the rate of magma supply. The goal is to find combinations of “unknown” parameters that allow us to match those parameters that are known.

Computer models of Kilauea are being used in two ways at HVO. The first is to assess what might be happening in places that we can’t directly observe. For example, the ground deformation at Kilauea’s summit suggests that there is a magma chamber that is inflating and deflating beneath the east margin of Halema`uma`u Crater. The computer model can be used to assess the depth of that magma chamber, which we have found to be about 1.5 km (1 mi) deep.

The second use is to forecast what might happen in the future. For example, if more magma was pushed into the magma chamber beneath Halema`uma`u, what volcanic activity might result? Would there be an eruption at the summit? Would there be a change in the eruption on the east rift zone?

As noted statistician George Box observed, however, “Essentially, all models are wrong, but some are useful.” While models are an important way of investigating volcanoes, we know they are inherently wrong, because they are based on assumptions and simplifications. No model will ever be “right,” but we can learn things from these models.

We’ll never reproduce the natural system perfectly or completely, but computer models provide a means of investigating these systems so we can learn more about how they work. The feedback between the models we test in Hawai`i and the activity that we observe at Kilauea and Mauna Loa allows us to refine the models and learn more and more about how volcanoes work.

Although scientific models have been in use for as long as there have been scientific observations, our new “supermodels” not only look good, but they represent a significant advance in the way we understand and interpret the data.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

 

Quantcast