Geodetic Monitoring at Axial Seamount Since its 2015 Eruption Reveals a Waning Magma Supply and Tightly Linked Rates of Deformation and Seismicity

Abstract

Axial Seamount is a basaltic hot spot volcano with a summit caldera at a depth of ˜1500 m below sea level, superimposed on the Juan de Fuca spreading ridge, giving it a robust and continuous magma supply. Axial erupted in 1998Axial erupted in , 2011Axial erupted in , and 2015 is monitored by a cabled network of instruments including bottom pressure recorders and seismometers. Since its last eruption, Axial has re-inflated to 85-90% of its pre-eruption level. During that time, we have identified eight discrete, short-term deflation events of 1-4 cm over 1-3 weeks that occurred quasi-periodically, about every 4-6 months between August 2016 and May 2019. During each short-term deflation event, the rate of earthquakes dropped abruptly to low levels, and then did not return to higher levels until reinflation had resumed and returned near its previous high. The long-term geodetic monitoring record suggests that the rate of magma supply has varied by an order of magnitude over decadal time scales. There was a surge in magma supply between 2011-2015, causing those two eruptions to be closely spaced in time and the supply rate has been waning since then. This waning supply has implications for eruption forecasting and the next eruption at Axial still appears to be 4-9 years away. We also show that the number of earthquakes per unit of uplift has increased exponentially with total uplift since the 2015 eruption, a pattern consistent with a mechanical model of cumulative rock damage leading to bulk failure during magma accumulation between eruptions.

Compartmentalization of Axial Seamount's magma reservoir inferred by analytical and numerical deformation modeling with realistic geometry

“…Our study models the observed inflation from 2016-2020, during a time when the magma supply rate was initially high, but then waned with time following the 2015 eruption (Chadwick et al, 2022). Given that the magma supply rate is estimated to have varied from >0.1 km 3 /year to <0.01 km 3 /year during that time period (Chadwick et al, 2022), our volume change estimates are reasonable. Mullet & Segall (2022) demonstrated that as the melt fraction of mush in a magma reservoir increases, the deformation caused by a mush-dominated magma storage system is increasingly driven by the overall shape of the mush body, instead of any pressurized melt lens within the mush.…”

“…The total volume of the shallow magma storage system beneath Axial was estimated by Arnulf et al (2014) to be 18-30 km 3 and the modeled co-eruptive volume change associated with previous eruptions has been estimated to vary between 0.147 -0.206 km 3 using analytical model source depths of 3-3.8 km (Chadwick et al, 1999(Chadwick et al, , 2012Hefner et al, 2020;. Our study models the observed inflation from 2016-2020, during a time when the magma supply rate was initially high, but then waned with time following the 2015 eruption (Chadwick et al, 2022). Given that the magma supply rate is estimated to have varied from >0.1 km 3 /year to <0.01 km 3 /year during that time period (Chadwick et al, 2022), our volume change estimates are reasonable.…”

“…Since MPR measurements were made in 2015 and 2017 (but not in 2016), we estimated the uplift values in 2016 at the MPR benchmarks by interpolating between the 2015 and 2017 MPR measurements assuming a linear deformation rate. The BPR record shows that deformation at the center of the caldera during this time period was not entirely linear (Chadwick et al, 2022). The benchmark at the center of the caldera had uplifted by 55 cm from mid-2015 to mid-2016, about 10 cm shallower in summer of 2016 than a linear interpolation would predict (Figure S1 in Supplementary Material).…”

“…BPRs were deployed at Axial's summit caldera in 1998 when Axial's first observed eruption occurred (Chadwick et al, 2013;Dziak & Fox, 1999;Embley et al, 1999;Fox, 1999;Fox et al, 2001). After a two-year gap in coverage, the deformation time series resumed in 2000 with an array of seafloor benchmarks and the time series has been continuous through the present (Figure 1; Chadwick et al, 2006Chadwick et al, , 2012Chadwick et al, , 2022Nooner & Chadwick, 2009. Since 2000, BPR measurements have been supplemented by measurements from mobile pressure recorders (MPRs), which are used in campaign-style surveys at seafloor benchmarks with a remotely operated vehicle (ROV) every 1-2 years to correct for the BPRs' long-term drift where the two are co-located (Chadwick et al, 2006).…”

“…Although viscoelastic effects and the presence of magma mush would likely not significantly impact the spatial distribution of modeled pressure changes in our results, it could impact volume change estimates due to magma compressibility. Modeling viscoelastic effects at Axial would be more strongly relevant to the temporal component of the deformation time series, for example to test hypotheses regarding short-term deflation events proposed by Chadwick et al, (2022).…”