Imaging the complex geometry of a magma reservoir using FEM-based linear inverse modeling of InSAR data: application to Rabaul Caldera, Papua New Guinea

Abstract

Summary We test an innovative inversion scheme using Green's functions from an array of pressure sources embedded in finite-element method (FEM) models to image, without assuming an a-priori geometry, the composite and complex shape of a volcano deformation source. We invert interferometric synthetic aperture radar (InSAR) data to estimate the pressurization and shape of the magma reservoir of Rabaul caldera, Papua New Guinea. The results image the extended shallow magmatic system responsible for a broad and long-term subsidence of the caldera between 2007 February and 2010 December. Elastic FEM solutions are integrated into the regularized linear inversion of InSAR data of volcano surface displacements in order to obtain a 3-D image of the source of deformation. The Green's function matrix is constructed from a library of forward line-of-sight displacement solutions for a grid of cubic elementary deformation sources. Each source is sequentially generated by removing the corresponding cubic elements from a common meshed domain and simulating the injection of a fluid mass flux into the cavity, which results in a pressurization and volumetric change of the fluid-filled cavity. The use of a single mesh for the generation of all FEM models avoids the computationally expensive process of non-linear inversion and remeshing a variable geometry domain. Without assuming an a-priori source geometry other than the configuration of the 3-D grid that generates the library of Green's functions, the geodetic data dictate the geometry of the magma reservoir as a 3-D distribution of pressure (or flux of magma) within the source array. The inversion of InSAR data of Rabaul caldera shows a distribution of interconnected sources forming an amorphous, shallow magmatic system elongated under two opposite sides of the caldera. The marginal areas at the sides of the imaged magmatic system are the possible feeding reservoirs of the ongoing Tavurvur volcano eruption of andesitic products on the east side and of the past Vulcan volcano eruptions of more evolved materials on the west side. The interconnection and spatial distributions of sources correspond to the petrography of the volcanic products described in the literature and to the dynamics of the single and twin eruptions that characterize the caldera. The ability to image the complex geometry of deformation sources in both space and time can improve our ability to monitor active volcanoes, widen our understanding of the dynamics of active volcanic systems and improve the predictions of eruptions.

Coeval Uplift and Subsidence Reveal Magma Recharging Near Rome (Italy)

“…However, the sensitivity of the surface deformations to elastic heterogeneities of the local crust has been studied and documented (e.g., Manconi et al, ; Masterlark, ). Recently, the petrography of volcanic products was found to impact on the interconnection and spatial distributions of volcanic sources retrieved by InSAR data at Rabaul caldera (Papua New Guinea), with direct implications on the dynamics of the caldera (Ronchin et al, ). Tomographic studies evidenced low‐velocity zones between 5 km and 16 km below the area of maximum uplift at CA (Cimini et al, ).…”

A Simplified Approach to Operational InSAR Monitoring of Volcano Deformation in Low- and Middle-Income Countries: Case Study of Rabaul Caldera, Papua New Guinea

“…We conducted an independent InSAR analysis for the Rabaul Caldera of the same ascending-pass ALOS-PALSAR Fine-beam L-band data used by Ronchin et al (2017) for the slightly extended period of 27 February 2007 to 10 March 2011 (21 images in total). The temporal coverage of the ALOS-PALSAR data used is highlighted in Figure 3.…”

“…InSAR was already applied to the measurement of surface deformation at the Rabaul Caldera: Ronchin et al (2017) conducted Persistent Scatterer InSAR analysis of ascending-pass ALOS-PALSAR data acquired between 27 February 2007 and 10 December 2010. Their results showed non-linear line of sight (LOS) displacement signals, localized to the Matupit and Vulcan peninsula regions, that were interpreted as subsidence indicative of the deflation of the shallow caldera magma body.…”

“…Their results showed non-linear line of sight (LOS) displacement signals, localized to the Matupit and Vulcan peninsula regions, that were interpreted as subsidence indicative of the deflation of the shallow caldera magma body. No inflation signal during the monitored time period was described by Ronchin et al (2017). The mean displacement (i.e., the FIGURE 3 | Record of absolute height change (with respect to the WGS84 Ellipsoid) at Matupit derived from RVO's daily processing of GNSS observations at station "SDA" since January 2000.…”

“…Within this construct, the InSAR data constrained the shape of a large, broadly cuboidal magmatic source between 3 and 4.5 km depth. A total magma volume of ∼1 × 10 8 m 3 was drained from this body during Tavuvur eruptions during the observation period (Ronchin et al, 2017). Despite the ability of this method to define a complex geometry for the shallow caldera magma body within a heterogeneous domain consisting of realistic earth properties, it is unlikely to be a useful approach for operational volcano monitoring in PNG due to the complexity and computational expense of the modelling approach.…”