Periodic Reporting for period 1 - 3DWISE (3D Full Waveform Inversion on seismic data at the East Pacific Rise)
Periodo di rendicontazione: 2015-09-01 al 2017-08-31
The mid-ocean ridge (MOR) system is the longest volcanic mountain chain that wraps around the Globe along which two plates have been separating and oceanic crust has been forming by highly active magmatic processes operating within and at the base of thus formed crust. It is believed that there is a strong link between these magmatic processes known at large spatial scales (1-100 km) and relatively localized (0.001-1 km) tectonic/volcanic/hydrothermal/biological processes, and their expression onto the seafloor. However, the characteristics of the link are largely unknown. The origin of the most of the unknowns is embedded in poorly defined properties of the upper crust (its structure, P and S wave velocities, porosity, density, etc.), where the link is hosted and its activity is taking place. To map these properties at an unprecedented resolution scale and unveil the existing link, we apply advanced three-dimensional (3D) full waveform inversion (FWI) techniques, which have been recently developed by industry, to a unique 3D seismic reflection data acquired at the East Pacific Rise (EPR). This portion of a MOR system is characterized by a prolific hydrothermal and magmatic activity and multidisciplinary time-series measurements, offering interdisciplinary approach to solve some of the fundamental problems in Earth sciences.
The overall goal of the 3DWISE project is to obtain high-resolution, 3D geophysical images of the upper oceanic crust applying a recently developed FWI technique for high-resolution P- and S-wave analysis to a 3D seismic dataset that was collected along the EPR (Figure 1).
The overall goal of the 3DWISE project is to obtain high-resolution, 3D geophysical images of the upper oceanic crust applying a recently developed FWI technique for high-resolution P- and S-wave analysis to a 3D seismic dataset that was collected along the EPR (Figure 1).
The work conducted during the 3DWISE project represents the first study of this kind done in 3D for the region 30x40 km2. Prior to the main task related to 3D work, extensive tests were done on a single 2D line collected along the ridge axis, which resulted in important observations. The techniques learned during the 2D work were then applied to conduct the 3D work. The entire cross-axis dataset, collected during the survey was then processed in 3D and 3D acoustic and elastic FWI was performed.
First, the results obtained from the 3-D work enable mapping of a contact between basaltic layer (layer 2A) and dikes (layer 2B) that together make the upper oceanic crust. In the velocity models, the layer 2A/2B boundary is characterized by a velocity gradient, which is attributed to change in porosity. The geologic nature of the gradient is debated, with the two prevailing explanations: lithological contact between basalts and dikes, or alteration front due to hydrothermal circulation. In addition, 2-D seismic sections suggested rapid thickening of the topmost layer within a few km from the ridge axis. Due to limited information on the upper crustal velocities it has been unclear if this observation is due to physical thickening of the extrusive layer or it is a result of downward propagating, hydrothermally driven, cracking front. The layer 2A/2B boundary is clearly identified in the resulting model as the base of high velocity gradient and can be followed throughout the entire area included in the inversion; consistency in character of the gradient zone and distinct velocity anomaly near active hydrothermal discharge zones, where the most of the alteration is expected to take place, argue that this boundary is predominantly lithological and that the layer 2A thickening is due to emplacement of lava off the innermost axial zone. The transition from thin (150-200 m) to thick (300-550 m) layer 2A occurs within a narrow band around the ridge axis (0.5-2.5 km). This band is wider between 9º48-53’, and highly asymmetric, with almost vertical side on the Pacific and gentle dipping side on the Cocos Plate, terminating at the contact with ridge parallel, inward facing faults. Beyond the faults, layer 2A attains almost constant thickness. By combining the available observables and results of our analyses we suggest that the emplacement of extrusives, variation in their thickness, and rate of dike subsidence are predominantly controlled by tectono-magmatic features and processes operating near the ridge axis.
In addition to exciting scientific results the 3DWISE project provides us with important insights on methodology that is used and outline recommendations when conducting work using wave-equations techniques: a) re-datuming (or downward continuation) should not be considered as a standard tool of data processing prior to the application of 3-D, acoustic and elastic FWI; b) if the elastic effect is present in the data it is very important to take it into account by conducting elastic inversion.
Second, the work along the ridge axis helped us to characterize the nature of the zero-age oceanic crust and look for potential geophysical signatures of hydrothermal flow within the ridge axis plane. Whilst the spatial distribution of the fluid discharge through the vent orifices is well documented, the distribution of fluid recharge zones along divergent plate boundaries and fluid flow pattern within the oceanic crust are still elusive. In the velocity models, region north of 9º45′N is represented by several low velocity perturbations, each >2 km wide. We interpret them as fluid flow pathways: the ones underlying vent clusters we relate to the presence of hydrothermal fluid discharge zones, whereas the ones collocated with higher-order tectonic discontinuities, we attribute to recharge zones. At depth of 1.5 km below seafloor, a signal from an axial magma lens (AML), capped by higher velocity conductive layer is imaged, creating ideal conditions for building up a hydrothermal plumbing system along the ridge axis plane. Within the eruption area three complete hydrothermal cells are identified, exhibiting important complexities in geometry of the channels with a low-angle, southward dip. Although prominent upper-crustal low velocity anomalies are identified in the vicinity of third-order discontinuities at 9º20′ and 9º37′N, the underlying crystallized AML hinders focused hydrothermal circulation. The geophysical imprints imaged using advanced inversion techniques argue that at the EPR, the presence of high thermal regime and high crustal permeability act in symbiosis and are equally important for developing, maintaining and driving vigorous, high-temperature hydrothermal flow.
First, the results obtained from the 3-D work enable mapping of a contact between basaltic layer (layer 2A) and dikes (layer 2B) that together make the upper oceanic crust. In the velocity models, the layer 2A/2B boundary is characterized by a velocity gradient, which is attributed to change in porosity. The geologic nature of the gradient is debated, with the two prevailing explanations: lithological contact between basalts and dikes, or alteration front due to hydrothermal circulation. In addition, 2-D seismic sections suggested rapid thickening of the topmost layer within a few km from the ridge axis. Due to limited information on the upper crustal velocities it has been unclear if this observation is due to physical thickening of the extrusive layer or it is a result of downward propagating, hydrothermally driven, cracking front. The layer 2A/2B boundary is clearly identified in the resulting model as the base of high velocity gradient and can be followed throughout the entire area included in the inversion; consistency in character of the gradient zone and distinct velocity anomaly near active hydrothermal discharge zones, where the most of the alteration is expected to take place, argue that this boundary is predominantly lithological and that the layer 2A thickening is due to emplacement of lava off the innermost axial zone. The transition from thin (150-200 m) to thick (300-550 m) layer 2A occurs within a narrow band around the ridge axis (0.5-2.5 km). This band is wider between 9º48-53’, and highly asymmetric, with almost vertical side on the Pacific and gentle dipping side on the Cocos Plate, terminating at the contact with ridge parallel, inward facing faults. Beyond the faults, layer 2A attains almost constant thickness. By combining the available observables and results of our analyses we suggest that the emplacement of extrusives, variation in their thickness, and rate of dike subsidence are predominantly controlled by tectono-magmatic features and processes operating near the ridge axis.
In addition to exciting scientific results the 3DWISE project provides us with important insights on methodology that is used and outline recommendations when conducting work using wave-equations techniques: a) re-datuming (or downward continuation) should not be considered as a standard tool of data processing prior to the application of 3-D, acoustic and elastic FWI; b) if the elastic effect is present in the data it is very important to take it into account by conducting elastic inversion.
Second, the work along the ridge axis helped us to characterize the nature of the zero-age oceanic crust and look for potential geophysical signatures of hydrothermal flow within the ridge axis plane. Whilst the spatial distribution of the fluid discharge through the vent orifices is well documented, the distribution of fluid recharge zones along divergent plate boundaries and fluid flow pattern within the oceanic crust are still elusive. In the velocity models, region north of 9º45′N is represented by several low velocity perturbations, each >2 km wide. We interpret them as fluid flow pathways: the ones underlying vent clusters we relate to the presence of hydrothermal fluid discharge zones, whereas the ones collocated with higher-order tectonic discontinuities, we attribute to recharge zones. At depth of 1.5 km below seafloor, a signal from an axial magma lens (AML), capped by higher velocity conductive layer is imaged, creating ideal conditions for building up a hydrothermal plumbing system along the ridge axis plane. Within the eruption area three complete hydrothermal cells are identified, exhibiting important complexities in geometry of the channels with a low-angle, southward dip. Although prominent upper-crustal low velocity anomalies are identified in the vicinity of third-order discontinuities at 9º20′ and 9º37′N, the underlying crystallized AML hinders focused hydrothermal circulation. The geophysical imprints imaged using advanced inversion techniques argue that at the EPR, the presence of high thermal regime and high crustal permeability act in symbiosis and are equally important for developing, maintaining and driving vigorous, high-temperature hydrothermal flow.
Beyond the state of the art proposed by the 3DWISE advanced techniques of seismic data migration were also applied. A detailed mapping of different generations of lava flow and faults within the shallow portion of the upper crust and around the ridge axis, will be possible. Moreover, a signal of potential lava flow that has been emplaced onto the seafloor during several decades and different eruption events and help us to have better estimates on its volume. This will help us to finally quantify the interplay between faults and lava flows in construction of the oceanic crust.
Moreover, imaging of the deeper portion of the crust enabled us to capture signal from the Axial Magma Lens (AML) and map the AML properties in three dimensions at the unprecedented resolution. Mapping the architecture of the event will help us to obtain some of the extrinsic properties of the event in 3D, its width and length of each segment.
Moreover, imaging of the deeper portion of the crust enabled us to capture signal from the Axial Magma Lens (AML) and map the AML properties in three dimensions at the unprecedented resolution. Mapping the architecture of the event will help us to obtain some of the extrinsic properties of the event in 3D, its width and length of each segment.