Release of mineral-bound water prior to subduction tied to shallow seismogenic slip off Sumatra

Abstract

International audienceTrying to understand where major earthquakes and tsunamis might occur requires analysis of the sediments pouring into a subduction zone. Thick sediments were expected to limit earthquake and tsunami size in the Sumatran megathrust event in 2004, but the magnitude 9.2 earthquake defied expectations. Hüpers et al. analyzed sediments recovered from the Sumatran megathrust. They found evidence of sediment dehydration, which increased fault strength and allowed for the much larger earthquake to occur. Thus, models of other subduction zones, such as the Gulf of Alaska, may underestimate the maximum earthquake magnitude and tsunami risk

Sumatra Subduction Zone

“…A preliminary interpretation is that this freshening reflects the first in a series of dehydration reactions, and further progression of these reactions could increase fluid pressures and influence sediment strength and petrophysical properties. Preliminary analysis of this chemical anomaly and implications for the evolution of the input section and plate boundary fault formation and behavior are described by Hüpers et al (2017). Additional sediment geochemical analyses and modeling work will further constrain the roles of these reactions on freshening and how they evolve with burial.…”

“…We analyzed a total of 28 IW samples at a frequency of 1 or 2 samples per core from Cores 362-U1481A-2R through 38R (1149.70-1498.72 mbsf). Whereas there are similarities in the dehydration reactions that may lead to fluid production at depth (see Hüpers et al, 2017), the differences in sediment composition of lithologic Unit III at Site U1481 and Subunit IIIA at Site U1480 are reflected by a lack of a pronounced increase in B and Mn at Site U1481 at the horizon of observed fluid freshening, marked by low Cl (Figures F10, F11). Deeper in the section, basement was not reached at Site U1481, and thus we do not have a clear indication of the effect that diffusion of ions from (or into) the basement aquifer may have on the pore fluid geochemistry at depth at this site.…”

“…Natural gamma radiation, magnetic susceptibility (MSL = magnetic susceptibility logger, MSP = point magnetic susceptibility), P-wave velocity (discrete samples), and MAD porosity were measured as part of the standard physical properties core flow. Chloride and silica are examples of the routine pore fluid geochemical measurements (see also Hüpers et al, 2017). Stratigraphic ages from biostratigraphic and magnetostratigraphic analysis of core samples (see also McNeill et al, 2017b).…”

“…Collectively, these concentration changes point to dehydration reactions between 1250 and at least 1300 mbsf. Hüpers et al (2017) provide detailed analysis on the source and timing of the deep chloride anomaly and implications for seismogenesis. Deeper in the sediment section, an increase in chloride (Figure F10), strontium, and calcium and a decrease in magnesium indicate a contribution from basement fluid.…”

“…The onset of Nicobar Fan deposition at the drill sites (~9.5 Ma; Hüpers et al, 2017;McNeill et al, 2017a) is significantly younger than was anticipated precruise (~30-40 Ma; see Background and objectives in the Site U1480 chapter [McNeill et al, 2017b], based on previous regional analyses of Bengal-Nicobar Fan history and presumptions of gradual fan progradation).…”