scholarly journals Joint interpretation of geophysical data: Applying machine learning to the modeling of an evaporitic sequence in Villar de Cañas (Spain)

2021 ◽  
pp. 106126
Author(s):  
I. Marzan ◽  
D. Martí ◽  
A. Lobo ◽  
J. Alcalde ◽  
M. Ruiz ◽  
...  
2021 ◽  
Author(s):  
Kun Wang ◽  
Christopher Johnson ◽  
Kane Bennett ◽  
Paul Johnson

Abstract Data-driven machine-learning for predicting instantaneous and future fault-slip in laboratory experiments has recently progressed markedly due to large training data sets. In Earth however, earthquake interevent times range from 10's-100's of years and geophysical data typically exist for only a portion of an earthquake cycle. Sparse data presents a serious challenge to training machine learning models. Here we describe a transfer learning approach using numerical simulations to train a convolutional encoder-decoder that predicts fault-slip behavior in laboratory experiments. The model learns a mapping between acoustic emission histories and fault-slip from numerical simulations, and generalizes to produce accurate results using laboratory data. Notably slip-predictions markedly improve using the simulation-data trained-model and training the latent space using a portion of a single laboratory earthquake-cycle. The transfer learning results elucidate the potential of using models trained on numerical simulations and fine-tuned with small geophysical data sets for potential applications to faults in Earth.


2019 ◽  
Author(s):  
M.I. Shimelevich ◽  
Е.А. Obornev ◽  
I.E. Obornev ◽  
E.A. Rodionov

2021 ◽  
Author(s):  
Irina M. Artemieva

<p>The lithosphere is a thermal boundary layer atop mantle convection and a chemical boundary layer formed by mantle differentiation and melt extraction. The two boundary layers may everywhere have different thicknesses. Worldwide, the thicknesses of thermal and chemical boundary layers vary significantly, reflecting thermal and compositional heterogeneity of the lithospheric mantle.</p><p>Physical parameters determined by remote geophysical sensing (e.g. seismic velocities, density, electrical conductivity) are sensitive to both thermal and compositional heterogeneity. Thermal anomalies are usually thought to have stronger effect than compositional anomalies, especially at near-solidus temperatures when partial melting and anelastic effects become important. Therefore, geophysical studies of mantle compositional heterogeneity require independent constraints on the lithosphere thermal regime. The latter can be assessed by various methods, and I will present examples for continental lithosphere globally and regionally. Of particular interest is the thermal heterogeneity of the lithosphere in Greenland, with implications for the fate of the ice sheet and possible signature of Iceland hotspot track.</p><p>Compositional heterogeneity of lithospheric mantle at small scale is known from Nature's sampling, such as by mantle-derived xenoliths brought to the surface of stable Precambrian cratons by kimberlite-type magmatism. This situation is paradoxical since “stable” regions are not expected to be subject to any tectono-magmatic events at all. Kimberlite magmatism should lead to a significant thermo-chemical modification of the cratonic lithosphere, which otherwise is expected to have a unique thickness (>200 km) and unique composition (dry and depleted in basaltic components). Nevertheless, geochemical studies of mantle xenoliths provide the basis for many geophysical interpretations at large scale.</p><p>Magmatism-related thermo-chemical processes are reflected in the thermal, density, and seismic velocity structure of the cratonic lithosphere. Based on joint interpretation of geophysical data, I demonstrate the presence of significant lateral and vertical heterogeneity in the cratonic lithospheric mantle worldwide. This heterogeneity reflects the extent of lithosphere reworking by both regional-scale kimberlite-type magmatism (e.g. Kaapvaal, Siberia, Baltic and Canadian Shields) and large-scale tectono-magmatic processes, e.g. associated with LIPs and subduction systems such as in the Siberian and North China cratons. The results indicate that lithosphere chemical modification is caused primarily by mantle metasomatism where the upper extent may represent a mid-lithosphere discontinuity. An important conclusion is that the Nature’s sampling by kimberlite-hosted xenoliths is biased and therefore is non-representative of pristine cratonic mantle.</p><p>I also present examples for lithosphere thermo-chemical heterogeneity in tectonically young regions, with highlights from Antarctica, Iceland, North Atlantics, and the Arctic shelf. Joint interpretation of various geophysical data indicates that West Antarctica is not continental, as conventionally accepted, but represents a system of back-arc basins. In Europe and Siberia, an extremely high-density lithospheric mantle beneath deep sedimentary basins suggests the presence of eclogites in the mantle, which provide a mechanism for basin subsidence. In the North Atlantic Ocean, thermo-chemical heterogeneity of the upper mantle is interpreted by the presence of continental fragments, and the results of gravity modeling allow us to conclude that any mantle thermal anomaly around the Iceland hotspot, if it exists, is too weak to be reliably resolved by seismic methods.</p><p>https://stanford.academia.edu/IrinaArtemieva</p><p>www.lithosphere.info</p>


2020 ◽  
Author(s):  
Adrian S. Barfod* ◽  
Jakob Juul Larsen

<p>Exploring and studying the earth system is becoming increasingly important as the slow depletion of natural resources ensues. An important data source is geophysical data, collected worldwide. After gathering data, it goes through vigorous quality control, pre-processing, and inverse modelling procedures. Such procedures often have manual components, and require a trained geophysicist who understands the data, in order to translate it into useful information regarding the earth system. The sheer amounts of geophysical data collected today makes manual approaches impractical. Therefore, automating as much of the workflow related to geophysical data as possible, would allow novel opportunities such as fully automated geophysical monitoring systems, real-time modeling during data collection, larger geophysical data sets, etc.</p><p>Machine learning has been proposed as a tool for automating workflows related to geophysical data. The field of machine learning encompasses multiple tools, which can be applied in a wide range of geophysical workflows, such as pre-processing, inverse modeling, data exploration etc.</p><p>We present a study where machine learning is applied to automate the time domain induced polarization geophysical workflow. Such induced polarization data requires pre-processing, which is manual in nature. One of the pre-processing steps is that a trained geophysicist inspects the data, and removes so-called non-geologic signals, i.e. noise, which does not represent geological variance. Specifically, a real-world case from Grindsted Denmark is presented. Here, a time domain induced polarization survey was conducted containing seven profiles. Two lines were manually processed and used for supervised training of an artificial neural network. The neural net then automatically processed the remaining profiles of the survey, with satisfactory results. Afterwards, the processed data was inverted, yielding the induced polarization parameters respective to the Cole-Cole model. We discuss the limitations and optimization steps related to training such a classification network.</p>


Sign in / Sign up

Export Citation Format

Share Document