The Norrbotten region in northern Sweden, as a part of the Fennoscandian (Baltic) Shield, is one of the most active mining areas in Europe. It is acknowledged that geological structures such as faults, shear zones, and associated fracture systems play a key role in providing a physical pathway connecting metal sources and the sites of mineral precipitation. In particular, several magnetotelluric (MT) surveys around the world have revealed a spatial relationship and correlation between mineral occurrences and regional scale conductive structures. Previous lithospheric studies in the Fennoscandian Shield have indicated enhanced conductivities in the Norrbotten region, and they served as a basis for the proposal to test the hypothesis of this genetic connection. Therefore, the current project was initiated with the aim of obtaining more detailed information on crustal and upper mantle structures in the Paleoproterozoic Norrbotten ore province by utilising a combination of regional geophysical data including, MT and potential field data. Potential field data (airborne magnetic, and ground gravity) are extensively used as the first choice in geophysical/geological mapping due to good areal coverage, whereas MT surveys are typically designed to answer more specific geological and geophysical questions. In order to improve utilisation of the geophysical methods, quantitative analyses and integration approaches have been developed and applied to the geophysical data.

This thesis covers four topics: (i) a regional-scale study of Norrbotten County using 3D inversion of newly measured MT data to reveal crustal geoelectric structures, (ii) integration of the resistivity model with regional-scale magnetic susceptibility and density models using an unsupervised classification approach to extract and analyse the possible correlation between modelled petrophysical properties and provide a geophysical domain classification as input to geological interpretations, (iii) identification of a regional-scale tectonic feature in northern Sweden by processing and analysing the potential field data, (iv) the development and evaluation of the feasibility of a classification approach designed to identify patterns in potential field data that are indicative of the type of geological environment.

The conductivity model of inverted new MT data reveals the presence of strong crustal electrical conductors. The conductance of thousands of Siemens within a generally resistive crust is modelled. Some of the conductors in the model have near-surface expressions and are spatially correlated with the location of known mineralisation. A significant part of middle crust conductors is elongated in a direction that coincides with parts of major deformation zones. The derived conductivity model provides a new insight into the tectonic unit boundaries in the area.

The magnetic susceptibility and density models were subsequently integrated with the results obtained from the MT data. The integration approach is based on a joint analysis of the three modelled physical properties by using an unsupervised classification approach referred to as Self-Organising Maps (SOM). The depth variations of the properties were included in the classification. The obtained domain classification is discussed with respect to the previously mapped/interpreted geological units and compositions. Some discrepancies between existing geological maps and the domain classification are noted for some areas. The discrepancies may partly be related to the fact that the surface geological features are compared with geophysical models that also include information at depth.

A detailed analysis of high-resolution potential field data in northern Sweden led to the discovery of a regional-scale strike-slip fault, which we refer to as the Norrbotten Mega Fault (NMF). The presence of a fault cutting through the entire area in a roughly N5E direction with horizontal displacement of 51.2 km can be traced from Karesuando at the Swedish-Finnish border in the north to the Archaean-Proterozoic boundary, which is marked by the Luleå-Jokkmokk Zone roughly 250 km towards the south. Altogether, the length, apparent displacement, and straight character of the proposed fault suggest that it represents a late-orogenic, brittle fault. The initial identification of the location of the fault was primarily based on visual inspection of potential field data. The estimation of the displacement and validity of the proposed NMF is supported by analyses of several higher-order spatial derivatives of the potential field data. The analyses included application of a neural network SOM classification and a visualisation approach with subsequent matching of well-defined anomalies and anomaly patterns.

The application of SOM for pattern recognition and classification were developed further with respect to analysis of potential field data. This involved various experiments on data feature definitions and extraction. Data features that capture information about spatial variations, both depth-wise and horizontally, are used as input to the SOM for domain classification. The feasibility of the methodology and the limitations are discussed with respect to the results derived from a synthetic model. The proposed approach is furthermore applied to measured data from the entire Norrbotten area.

The models derived from the geophysical data, as well as the new integration and classification approaches presented in this study, may be applied as a guidance for future exploration and investigations in the area, such as for planning or expanding geophysical surveys in high prospective areas, mapping and interpretation of the geology, mapping the mineral prospectivity or other prediction tasks, or used as a guidance for follow-up on geological and geochemical work.

Potential field data in databases of the Geological Survey of Sweden (SGU) combined with newly acquired broadband magnetotelluric data are used to map and interpret geological units and structures of a 200 km by 250 km area in the Paleoproterozoic Norrbotten ore province (northern Sweden, latitudes 66°–68.5° and longitudes 19°–24°). In order to achieve this, a new approach is proposed with respect to extracting and analysing the possible correlation between modelled physical properties as well as their patterns with respect to depth variation within the crust. In this study, we propose the use of a neural net self-organising map procedure (SOM) for simplification, data reduction, and domain classification of the models derived from independent 3-D geophysical inversion of magnetotelluric, gravity, and magnetic data. The crustal model of the electrical conductivity structure was obtained from previous 3-D inversion of the magnetotelluric data. Processing and 3-D inversion of the regional magnetic and gravity field data were performed using an open-source object-oriented code called SimPEG. The input data to the SOM analysis contain resistivity, magnetic susceptibility, and density model values within the Norrbotten area for some selected depth levels of the entire crust. The domain classification is discussed with respect to the geological boundaries and composition of the crust. Consistency between model domain classification and geological boundaries is observed in general but an apparent discrepancy is noted for some areas. The reason for the apparent discrepancy is likely related to that most geological boundaries represent surface features whereas the geophysical data includes information at depth.

We present a fairly large, Potential Idiomatic Expression (PIE) dataset for Natural Language Processing (NLP) in English. The challenges with NLP systems with regards to tasks such as Machine Translation (MT), word sense disambiguation (WSD) and information retrieval make it imperative to have a labelled idioms dataset with classes such as it is in this work. To the best of the authors’ knowledge, this is the first idioms corpus with classes of idioms beyond the literal and the general idioms classification. Inparticular, the following classes are labelled in the dataset: metaphor, simile, euphemism, parallelism, personification, oxymoron, paradox, hyperbole, irony and literal. We obtain an overall inter-annotator agreement (IAA) score, between two independent annotators, of 88.89%. Many past efforts have been limited in the corpus size and classes of samples but this dataset contains over 20,100 samples with almost 1,200 cases of idioms (with their meanings) from 10 classes (or senses). The corpus may also be extended by researchers to meet specific needs. The corpus has part of speech (PoS) tagging from the NLTK library. Classification experiments performed on the corpus to obtain a baseline and comparison among three common models, including the state-of-the-art (SoTA) BERT model, give good results. We also make publicly available the corpus and the relevant codes for working with it for NLP tasks.

The Fennoscandian Shield as a part of a large Precambrian basement area is located in northern Europe and hosts economically important mineral deposits including base metals and precious metals. Regional geophysical data such as potential field and magnetotelluric data in combination with other geoscientific data contain information of importance for an understanding of the crustal and upper mantle structure. Knowledge about regional-scale structures is important for an optimized search for mineralisation.

In order to investigate in more detail the spatial distribution of regional electrically conductive structures and near-surface mineral deposits, complementary magnetotelluric measurements have been done within the Precambrian Shield in the north-eastern part of the Norrbotten ore province. The potential field data provided by the Geological Survey of Sweden have been included in the current study.

Processing of magnetotelluric data was performed using a robust multi-remote reference technique. The dimensionality analysis of the phase tensors indicates complex 3D structures in the area. A 3D crustal model of the electrical conductivity structure was derived based on 3D inversion of the data using the ModEM code. The final inversion 3D resistivity model revealed the presence of strong crustal conductors with the conductance of more than 3000 S at depth of tens of kilometres within a generally resistive crust. A significant part of the middle crust conductors is elongated in directions that coincide with major ductile deformation zones that have been mapped from airborne magnetic data and geological fieldwork. Some of these conductors have near-surface expression where they spatially correlate with the locations of known mineralisation.

Processing and 3D inversion of the regional magnetic and gravity field data were performed, and the structural information derived from these data by using an open-source object-oriented package code written in Python called SimPEG. In this study, a new approach is proposed to extract and analyse the correlation between the modelled physical properties and for domain classification. For this, a neural net Self-Organizing Map procedure (SOM) was used for data reduction and simplification. The input data to the SOM analysis contain resistivity, magnetic susceptibility, and density model values for some selected depth levels. The domain classification is discussed with respect to geological boundaries and composition. The classification is furthermore applied for prediction of favourable areas for mineralisation.

Based on visual inspection of processed regional gravity and magnetic field data and a SOM analysis performed on higher-order derivatives of the magnetic data, an interpretation of a sinistral fault with 52 km offset is proposed. The fault is oriented N10E and can be traced 250 km from Karesuando at the Swedish-Finish border southwards to the Archaean-Proterozoic boundary marked by the Luleå-Jokkmokk Zone.

Magnetotelluric data from 104 new stations in northern Sweden together with previously available MaSca project data are analysed with respect to a hypothesised genetic relation between mineral occurrences and commonly reported deep crustal electrically conductive structures in Precambrian Shield areas. A total of 165 stations from the north-eastern part of the Norrbotten ore province are modelled from an area characterised as having an Achaean basement with later reworking during the Paleoproterozoic. The data cover an area of 200 × 250 km² and provide a significant improvement of spatial resolution compared to previously available data. The derived 3D model reveals the presence of strong crustal conductors with conductance of more than 3000 S at depth of tens of kilometres within a generally resistive crust. A subdivision into a central more conductive domain bounded towards north and south by resistive domains is revealed. The boundaries between these three domains are oriented NW-SE. The southern boundary coincides roughly with the boundary between Archaean and Paleoproterozoic crust defined by ε_Nd data. The northern boundary coincides with an interpreted 2.4–2.3 Ga coastline of a marine rift basin, which is spatially defined by the occurrence of the Kovo group rocks. The location of major conductors within the central domain correlates with the presence of mapped Paleoproterozoic metasupracrustal belts. A significant part of the middle crust conductors is elongated in directions that coincide with major deformation zones that have been mapped from airborne magnetic data and geological fieldwork. Some of these conductors have near-surface expression where they spatially correlate with the location of known mineralisation. In these cases, the conductive structures may mark the pathways for metal-bearing hydrothermal fluids. Interpretations involving enhanced quantities of magnetite or graphite in the middle crust are discussed but the magnetotelluric data alone are not conclusive with respect to cause of observed high conductivities.

Broadband magnetotelluric data were recorded at 104 sites between 2015-2018 in northern Sweden to image the geoelectrical upper and lower crustal structures. Data processing was performed using a robust multi-remote reference technique. The dimensionality analysis of the phase tensors indicate complex 3D structures in the area. A 3D crustal model of the electrical conductivity structure was derived based on 3D inversion of the data. Processing of regional potential field data was performed and structural information derived from these data were compared with the 3D conductivity model.

Locations of known mineralizations are compared to the regional geophysical data in order to investigate how the regional geophysical data can be used for better informed mineral exploration. The analyses indicate that regional geophysical can provide very useful information with respect to the prospectivity of different areas.