Observations and models of ground deformation from the PLUTONS Project: Lazufre and Uturuncu, Central Andes

The Central Andes Volcanic Zone (CVZ, 14°-28°S) is one of three distinct arc segments the Nazca-South America subduction system. In comparison to the Northern and Southern segments, the CVZ contains approximately 40% of volcanoes active during the Holocene, but less than 20% of documented eruptions. It is therefore surprising that synoptic geodetic observations since 1992 have so far revealed half of the 20 known uplifting volcanoes in the Andes are in the CVZ. Furthermore, an especially high concentration of Miocene ignimbrite deposits (> 10,000 km^3) suggests that in the past large volumes of eruptible magma traversed the crust in this region. We utilize geodetic modeling to address the following questions: What are physically plausible depths, geometries, volumes, and transport mechanisms of intrusions? What are the conditions for plutonism versus volcanism in the CVZ?

Our modeling efforts are focused on two of the spatially largest (>2,000 km^2) volcanic uplift events observed globally (Lazufre and Uturuncu). We present a synthesis of InSAR, continuous GPS, and campaign GPS (collected from a small network of sites around Lazurfre in Nov. 2011 and March 2014). New InSAR processing of Envisat ScanSAR (09/2003 - 11/2009), ALOS (02/2007 - 02/2011), and TSX (04/2008 - 07/2014) confirm continued Lazufre uplift rates of approximately 3 cm/yr. Neither TSX data (06/2012 – 07/2014), nor two continuous GPS sites on and around Uturuncu show evidence of continued 1 cm/yr vertical motion since the sites were installed in April 2010.

Analytic elastic models of uplift suggest intrusions accumulate in reservoirs in the mid-to-upper crust at both volcanic centers. However, peripheral subsidence at Uturuncu and observations of an extensive low velocity zone motivate the exploration of alternative realistic models that consider the influence of a feeder reservoir in the lower crust and heterogeneous crustal structure. The dual-reservoir model provides a first-order estimate of magma ascent rates, but there is a tradeoff with crustal rheology. Finite element forward models and inverse models with FEM-generated Green’s functions demonstrate that both reservoir separation distance and seismically-defined heterogeneity are significant and effect components of surface displacement disproportionately.