GRACE GRAVITY INFERRED CRUSTAL DEFORMATION MODELS OF SOME MEGATHRUST EARTHQUAKES

Abstract

The mega-thrust offshore earthquakes leading to Tsunami generation led to loss of human lives, affected economies of several nations through destruction of property and infrastructure. However, these earthquakes have also provided necessary geophysical data, whose analysis led to a deeper understanding of episodic subduction processes involving large scale deformations at several depth levels from hypocenter to mean sea level. GRACE satellite with its global gravity mapping capability in space and time have provided an immense opportunity to study the episodic processes of subduction and associated deformations of geoid, crust and upper mantle. By considering the conventional gravity modeling of subduction of oceanic plate below continental plate at oceanic trench region, our co-seismic GRACE gravity anomalies are modeled using unit vertical pyramid model. The co-seismic gravity signals are arrived at by differencing post from pre – earthquake GRACE gravity data at a mean satellite height of 500 km. The resulting data is filtered by a carefully selected band-pass filter after a systematic search over the entire range of spherical harmonic spectrum of 90 degrees and orders. A carefully constrained 3-D density distribution and relevant multi-layered thrust fault spanning crust and mantle in hypocenter regions of offshore mega-earthquakes like Sumatra Earthquake 2004, Chile Earthquake 2010, Japan Earthquake 2011 and Okhotsk Earthquake 2013 (Deep focus earthquake) have been considered for estimating the gravity responses, which matched with observed anomaly pattern (RMS errors are 6.26%, 5.80%, 6.59% and 8.7%). We have also independently derived crucial seismological parameters (rupture length, slip rate, seismic moment, momentum, earthquake energy distribution, force, differential pressure and work done) for these mega-earthquakes and they agree with the literature. Our co-seismic GRACE gravity modeling comprising an eleven layered 3-D megathrust fault with varying slip (𝜃1-𝜃10 ), in the case of shallow earthquakes like Sumatra Earthquake 2004, Chile Earthquake 2010 and Japan Earthquake 2011, and varying slip (𝜃1-𝜃15 ) for fifteen layered 3-D mega-thrust fault for deep focus Okhotsk Earthquake 2013 have honored co-seismic deformations from sea surface up to a hypocenter and beyond. The roles of seismic energy partitioning, frictional heat loss, spherical spreading energy loss and absorption losses ii in the earth medium are considered for gravity based estimates of slip rate. Co-seismic gravity data analysis inferred momentum at sea floor can be equated to an area pulse in motion led to deformation of ocean surface (geoid) in the range of minimum to maximum, which can serve the Tsunami generation model(s). Further, this has helped to interpret the pressure (areal) pulse created due to co-seismic deformation of ocean floors leading to deformation of geoid and tsunami generation in case of Sumatra 2004, Chile 2010 and Japan 2011mega earthquakes. However, the physical mechanism of deep earthquakes (depth>300km) remains a complex puzzle, partly because their rupture dimensions are difficult to estimate due to their low aftershock activity and absence of geodetic or surface rupture observations. Satellite gravity data analysis and interpretation provided a unique opportunity in understanding the rupture processes of such deep focus earthquakes to a first order approximation by a rigid body motion within brittle earth medium. Our gravity inferred seismological parameters may serve as additional independent constraints to traditional seismological studies. Thus, our traditional gravity modeling for co-seismic gravity signals based on brittle mechanical model of affected earth medium have helped in 3-D mapping of the deformed volumes of geoid, ocean floor, crust and mantle (up to hypocenter and beyond) in a novel manner contrary to more sophisticated earth media extending from elastic to visco-elastic involving more intricate relaxation models in seismological literature. Further, our independent estimates of seismological parameters for the attempted four mega-earthquakes and their successful validation with literature gives us a hope that our modeling efforts could be undertaken as a preliminary step prior to attempting more intricate deformation models for coseismic and post-seismic satellite gravity response analyses. In such an effort, the ambiguity in our gravity models can be reduced further with more refined input data sets like fortnightly to daily based spherical harmonic solutions in open file platforms for satellite gravity.