Cruise informatiom

R/V Thompson | ROV Jason and AUV Sentry | Seattle-Seattle, August 14-29

Forecasting Eruptions

Tuesday August 25, 2015 

Overcast skies with calm winds (less than 15 knots) and relatively calm seas (swells of 6-8 ft).

What’s happening today? 
 ROV Jason continues with the last stages of the pressure dive and has resumed collecting fluid samples where hydrothermal vents are near pressure benchmarks. AUV Sentry continued mapping the caldera, then was recovered at this afternoon.

 Forecasting eruptions
 By Bill Chadwick

Graphic showing magma intruding into the volcano causing inflation.
Volcanoes like Axial Seamount inflate and deflate like a big balloon in response to magma going into or out of a reservoir located beneath the summit caldera. Between eruptions the volcano slowly inflates (and the seafloor gradually rises by 10s of cm/yr) as magma is injected into the reservoir from below. Then during eruptions, the volcano rapidly deflates (and the seafloor drops by 2-3 m over a few days) as magma is removed from the reservoir to erupt as lava on the seafloor.

Graphic showing magma erupting to the surface of the volcano, causing deflation.
The pressure measurements we are making during the current Jason dive measure this vertical movement at many locations on the volcano and they are useful in several ways. They allow us to estimate the depth, size, and shape of the magma reservoir that is pushing the seafloor up and down from below. For example, from the depths determined by the pressure measurements we know the magma reservoir is located a few kilometers (~1 mile) below the seafloor and is centered on the caldera. We can also use the measurements to estimate the supply rate of magma into (or out of) the reservoir. For example, we know that the magma supply at Axial has been continuous for the last 18 years, but the rate of that supply has changed with time. After the 2011 eruption the supply rate was 4 times larger than before the 2011 eruption. That is why the time interval between the 2011-2015 eruptions was shorter (4 yrs) compared with the previous interval (13 yrs between 1998-2011). The inflation and deflation measurements can also be used to anticipate when the next eruption might occur.

Plot of bottom pressure recorder (BPR) mobile pressure measurement data (MPR) with time, showing repeated pattern of rapid deflation during eruptions (drop of the seafloor), followed by slow inflation between eruptions (uplift of the seafloor). These data plus the MBARI AUV data allowed us to forecast an eruption in 2015.
To attempt to forecast eruptions at Axial, we’ve used the simple idea that the volcano should be ready to erupt again when its level of inflation reaches the same threshold it reached when the previous eruption was triggered. That idea is based on the concept that as the magma reservoir inflates with magma, the pressure inside gradually builds (like more and more air forced into a balloon). Eventually the pressure gets so high that magma forces its way out of the reservoir (like a balloon popping or springing a leak). That’s when an eruption happens. Of course, not all volcanoes are that “well behaved,” and their pattern of inflation and deflation is not necessarily repeatable or predictable from one eruption to the next. But we wondered if a submarine volcano like Axial might have a predictable pattern since the ocean crust is so thin and the magma “plumbing system” is relatively simple. Once we started measuring the pattern of inflation and deflation at Axial we decided to give it a try.

Same plot as above with deflation pressure data that was recorded during the 2015 eruption by an instrument on the Ocean Observatories Initiative Cabled Array network (in orange). We are collecting data during the Axial Seamount 2015 expedition that will more closely tie data recorded during the April 2015 eruption to the previous time-series. More information at:
Our first attempts were complicated by the fact that we started with a gap in the pressure data after the 1998 eruption and our early ROV-based pressure measurements had relative high errors as we learned how refine the technique. Despite those uncertainties, in 2006 we made a forecast that we thought Axial would erupt again “before 2014”. Our anticipated eruption occurred in 2011 – and even though the forecast window was pretty wide (several years), we were encouraged by this success. Since then, our measurement techniques have continued to improve, which allowed us to make a forecast with a narrower time window before the 2015 eruption. In September 2014, we issued a forecast that Axial would erupt again sometime in 2015, and that eruption occurred in late April. Another success! A critical piece of information to that successful forecast was repeat bathymetry by our colleagues at MBARI with their AUV that showed that the rate of inflation we had measured with our pressure sensors in 2013 was continuing into 2014. Our hope is that by studying how eruptions can be forecast at Axial Seamount we will learn valuable lessons that can be applied to predicting volcanic eruptions on land.

To learn more about the 2015 Axial eruption, see the video from the earlier posting about the eruption.