A5067218711- Reservoir Engineering and Simulation Methods
- Geophysical and Geoelectrical Methods
- Geophysics and Gravity Measurements
- Planetary Science and Exploration
- Seismic Waves and Analysis
- Seismic Imaging and Inversion Techniques
- Advanced MEMS and NEMS Technologies
- Geophysics and Sensor Technology
- Geological Modeling and Analysis
University of Glasgow
2022-2023
City of Glasgow College
2022
Satellite gravimetry data from Mars offers a unique glimpse into the planet's interior structure. Combined with topography of surface, measurements gravitational field can be used to probe lateral density variations in upper layers. Due inherent degeneracies between effects anomalies mantle and crust on gravity, incomplete isostasy models, previous efforts for global gravity inversion decouple two planetary layers were unsuccessful. This study aims aid these by providing constraining...
Normalising flows is a novel generative neural network model, which can be applied to Bayesian parameter inference. When gravity inversion reformulated as probabilistic inference problem, stable results obtained that naturally incorporate the inherent uncertainties and noise from source background instrument. As opposed some standard methods, does not default single solution in an ill-posed but informs user about all possibilities are consistent with gravimetry survey of interest. It has...
This work is concerned with applying machine learning to Bayesian gravity inversion. While various frameworks have been explored for this application, these methods are not ideal due the associated long computation time and can become intractable a high-dimensional parameter space. For many applications, offer faster approach obtaining posterior approximations, while sacrificing accuracy. Normalising flows, which based on simple principles of change variables formula, recently focus...
Gravity inversion is a commonly applied data analysis technique in the field of geophysics. While machine learning methods have previously been explored for problem gravity inversion, these are deterministic approaches returning single solution deemed most appropriate by algorithm. The method presented here takes different approach, where reformulated as Bayesian parameter inference problem. Samples from posterior probability distribution source model parameters obtained via implementation...
<p>The measurement of tiny variations gravity over long time-scales or across the landscape has been interest for geophysicists and various industries since development first modern gravimeter. The manufacturing cost overall survey time required with commercial gravimeters, however, limit their potential application. MEMS gravimeter developed at University Glasgow, Wee-g, is a small form-factor, high-sensitivity relative under development, its low enabling to be used in...