Volatiles and Fluids in Subduction Zones

Advanced model (A2 data and other features ) (04/2007)

Our 3D density model of the Central American subduction zone has been further constrained. This model is based on combination of constraints from curvature analysis, Euler deconvolution and other constraining data including geological maps (covering the surface of the model), structural information from seismic profiles and earthquake hypocenters. In addition, new tomographic and seismic work from subproject A2 and A5 are included.

The sources of gravity anomalies have also been investigated using the Euler deconvolution method. From the theoretical point of view, this method allows estimation of depths to the sources of density anomalies in the lithosphere. Source point clusters, obtained at depths of 10 to 30 km, help to design a regional density model for Central America, Pacific and Caribbean lithosphere. Especially in the fore-arc region, the cluster correlates well with the top of the subducting plate. In the present analysis, only the best solutions have been selcted and most of the solutions are concentrated in the southern part of the study area. The reliability of the results in the other parts seems doubtful. New seismological data in Nicaragua and Costa Rica from the TUCAN group (G. Abers Boston University, USA & K. Fischer Brown University, USA) and the first seismic profile along the volcanic arc from the group of S. Holbrooke (University of Wyoming, USA) are essential, because they cover the unconstrained areas of the volcanic arc in the northern part of Costa Rica and the southern part of Nicaragua. Unfortunately these data are not yet available.

The model is roughly divided into two parts: the northern part where a normal subduction takes place and the southern part where the Cocos Ridge (track of the Galapagos hotspot with a thickened oceanic crust up to 20 km: Walther, 2002) collides with southern Costa Rica and subducts. In contrast to the previous publication (Stavenhagen, 1998), the angle of subduction is much steeper than expected, and this is well constrained by earthquake hypocenters (subproject A2). Preliminary results from the TUCAN group show only small changes in the subducting slab but major changes within the lithosphere. This corresponds to changes in gravity and offset of the volcanic arc. Preliminary results from the receiver function analysis show Moho depth very close to the one calculated using isostatic method (Vening-Meinesz-method). However, the results are tentative, and were presented during the conference in June 2007 in Heredia, Costa Rica (Workshop to Integrate Subduction Factory and Seismogenic Zone Studies in Central America).

The 3D density model of the crustal and lithospheric structure in Central America provides a base for further analysis and interpretation. It combines results from various methods and provides for the first time a synoptic picture of the investigated area down to the upper mantle. It will also serve as a base model for other subprojects to expand their results to 3D.

The observed serpentinisation in the Cocos Plate, offshore Nicaragua (Berhorst, 2006; subproject A5) was modelled as a 5 km thick layer with a density reduction of 0.1 Mg/m³ (serpentinisation of ~ 15%). The effect without serpentinisation is remarkably high (up to 30*10-5 m/s²). However, the modelling is based on relative gravity values. Therefore, precise answers whether a serpentinzed layer exists or not cannot be given from density modelling without other constraint.

The 3D model also forms the base for a more detailed investigation of the crustal structure associated with the volcanic arc in Costa Rica. Current small scale modelling by O. Lücke in the Central valley of Costa Rica gives new insights about the deep structure of the volcanic arc . Two major low densities crustal bodies are being modelled which are representative of volatile rich melt zones (i.e. magma chambers). The first major low density body (density 2.3 Mg/m³) corresponds to the Quaternary volcanic arch, and the second low density body (density 2.4 Mg/m³) corresponds to the Miocene-Pliocene Aguacate paleo-arc. A high density crustal body (3.3 Mg/m3) has been used to explain the gravity high on the back-arch Sarapiquí and Tortuguero plains and is interpreted as an uplifted block of oceanic crust basement

Due to publication issues, some pictures have not been published here. We appologize for this!