Lp-norm inversion constrained by geological information using neural network and full-tensor gravity gradiometry data
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Traditional inversion methodologies encounter substantial challenges in assimilating geological information, such as the spatial extent of geological structures and the hypothesized boundaries of anomalous bodies. We proposed a neural network-assisted framework, designed to integrate geological information by adaptively and flexibly delineating the model space for inversion utilizing full-tensor gravity gradiometry data. The model space can be enriched by incorporating diverse geological inputs, such as pre-existing geological models, lithological indicators obtained from drill cores, or surface field observations (e.g., outcrops). Through the random sampling of user-adjustable parameters within the formulated objective function, the framework accommodates both smooth and compact subsurface features, and captures the uncertainties of reconstructed models, thereby providing more dependable interpretations. We conducted both synthetic and field examples to illustrate the efficacy of the proposed framework.
Fields of Study & Subjects
Geophysical and Geoelectrical Methods
Geophysical Methods and Applications
Geophysics and Gravity Measurements
Publication details
- Journal / source
- Second International Meeting for Applied Geoscience & Energy
- Volume
- 1
- Pages
- 1587–1591
- Publication date
- 10 February 2026
- Language
- EN