Page 12 - GeoPRISMS_Newsletter_Spring2019_42_Neat
P. 12

a            Mixing events                                 b
                                                                         Normalized density
             0                                                     0

                                                                                        Edifice     Figure 3. Run-up to the
             3                                                     3
                                                                                                1999 eruption of Shishaldin
                                                                                                   volcano, simplified after
          Depth below summit (km)  12                       Eruption  12     Entrapment depth  Last storage depth  crystal clocks (green stars)
                                                                                                  Rasmussen et al. (2018c).
                                                                                                   (a) The timing of mixing
                                                                                                     events inferred from
                                                                                                 coincides with long-period
                                                                                                  (LP) and volcano-tectonic
                                                                                                  (VT) seismicity leading up
                                                                                                    determined from melt
                                                 Seismicity       15                             to the eruption. (b) Depths
            18                                                    18                              inclusions indicating that
                                                                                                  magmas coalesced at <3
                                                       LP                                         km depth (in the edifice).
                                                       VT                           Bin = 1 km
            21                                                    21
                                                                        Melt inclusion depths
                05/98  07/98   09/98  11/98   01/99  03/99   05/99
        We conducted a focused study of the 1999 eruption of Shishaldin  the maximum observed water contents, which minimizes the influence
        volcano (Rasmussen et al., 2018c), which is compelling for multiple  of diffusive leakage of water. Water contents are variable (~2-5 wt.%)
        reasons. This sub-Plinian eruption had an unusually long phase of  and correlate positively with geophysically determined magma storage
        seismic activity preceding the eruption, and, despite the 43 million m   depths, falling along the water-saturation curve (Fig. 4). The maximum
        of tephra ejected, InSAR data recorded no discernable eruption-related  water contents of melt inclusions often correlate with non-volatile
        deformation. We established a close temporal link between increased  trace elements, indicating diffusive leakage is not a major factor. Thus,
        seismicity, stress field changes inferred from shear-wave splitting  our data support a model in which intrinsically drier magmas (like
        analysis, and magma mixing recorded by crystal clocks, confirming   those that feed Shishaldin) degas and crystallize shallower than wet
        that precursory seismicity tracked the priming of the magmatic   magmas, resulting in shallower storage prior to eruption. So, what
        system for eruption by delivery of new magma (Fig. 3). Our study   controls primary water content?
        was the first to connect timescale information recorded in chemical
        zonation patterns in crystals with depth information recorded in melt   Now in the final leg of our pursuit to understand the slab-volcano
        inclusions, which we used along with geophysical data to interrogate   connection, we are focusing on the extent to which slab depth relates
        a previously enigmatic magmatic plumbing system. We found that a   to the composition of arc magmas (Rasmussen et al., 2018a). We have
        shallow magmatic system located largely within the edifice (<3 km   collected major, trace, and volatile element data in melt inclusions and
        below the summit) persists between eruptions. Prior to the 1999   bulk rock samples from the eight target volcanic centers in our corridor
        eruption, the shallow magmatic system was recharged with deep   (Fig. 1). These data exhibit systematic trends with slab depth (Fig. 5).
        (>20 km) magma. InSAR observations at Shishaldin are insensitive to
        deformation emanating from the shallow and deep parts of the system,   Figure 4. Relationship between the maximum water
        in part explaining the lack of an observed deformation signal. These   content of melt inclusions and geophysically determined
        results improve our understanding of eruption triggers and magmatic   magma storage depths, which are referenced to the
        plumbing systems. But this work raises the question, is the magmatic        volcano summits (Rasmussen et al., 2018b).
        system at Shishaldin unusually shallow?
        Broadening our scope, we evaluated magma storage depths
        throughout our corridor (Rasmussen et al., 2018b). Geophysical   2  Fisher
        constraints indicate these depths vary significantly (~2-8 km below                Okmok
        the edifice; Fig. 4). The cause of such variability is poorly understood.   Shishaldin
        Some have argued for the importance of intrinsic (e.g., buoyancy,   4                     Seguam
        viscosity) controls (Annen et al., 2006), while others have emphasized                     Cleveland
        the importance of extrinsic (e.g., crustal structure) controls (Chaussard   Depth (km)  6   Westdahl
        and Amelung, 2014). We investigated the influence of magmatic water   Fisher
        content, a key intrinsic variable, on magma storage depth. Water is   8     Makushin                 Akutan
        thought to be important because decompression-induced degassing
        during magma ascent results in an increase in melt viscosity and   10                           H2O saturation
        magma crystallinity, both promoting stalling. We estimated magmatic
        water content by measuring large suites of melt inclusions and taking   12
                                                                    1        2         3         4        5         6
                                                                                        H O (wt%)
        12  •  GeoPRISMS Newsletter  Issue No. 42  Spring 2019                            2
   7   8   9   10   11   12   13   14   15   16   17