Vesa NykänenGeological Survey of Finland, Rovaniemi, Finland · Digital products and services
Research Items (48)
Machine learning based classification methods are widely used in geoscience applications, including mineral prospectivity mapping. Typical characteristics of the data, such as small number of positive instances, imbalanced class distributions and lack of verified negative instances make ROC analysis and cross-validation natural choices for classifier evaluation. However, recent literature has identified two sources of bias, that can affect reliability of area under ROC curve estimation via cross-validation on spatial data. The pooling procedure performed by methods such as leave-one-out can introduce a substantial negative bias to results. At the same time, spatial dependencies leading to spatial autocorrelation can result in overoptimistic results, if not corrected for. In this work, we introduce the spatial leave-pair-out cross-validation method, that corrects for both of these biases simultaneously. The methodology is used to benchmark a number of classification methods on mineral prospectivity mapping data from the Central Lapland greenstone belt. The evaluation highlights the dangers of obtaining misleading results on spatial data and demonstrates how these problems can be avoided. Further, the results show the advantages of simple linear models for this classification task.
GIS based prospectivity mapping for different commodities at different scales is a common method to facilitate mineral exploration. The USGS three part method is another tool to assess the potential of undiscovered resources within a "permissive area". In this example we modeled the prospectivity of northern Finland for VMS deposits using knowledge-driven, fuzzy logic method and a combination of fuzzy logic and the USGS method. Data sets used for the modelling were selected based on diagnostic features of VMS deposits, and include regional till chemistry (5 elements), airborne geophysics, and ore showings data bases for VMS and related ore types, of the Geological Survey of Finland (GTK).
GIS based prospectivity mapping methods have is a cost-effective and time-saving method for outlining most prospective areas for selected deposit type and focusing exploration. In this case study a knowledge-driven prospectivity model for IOCG deposits in northern Finland is generated using a fuzzy logic method. Evidential data sets are selected based on the diagnostic features of the IOCG deposits outlined in literature. Derivatives of airborne geophysics, regional till geochemistry, geological maps, and ore showing data bases of the GTK comprise the data sets used in the model. The Final model is statistically validated using Receiver Operating Characteristics (ROC) method. The resulting mineral prospectivity map outlines the most favorable areas for IOCG deposits in northern Finland. ROC validation yield Area Under Curve (AUC) score of 0.963 indicating a very good fit of the model.
Finland is a traditional mining country and currently one of the most active regions for exploration and mining investments in Europe. Globally, Finland has been ranked in the top 5 in mining investment attractiveness since 2010. The key factors behind the success are the diversified mineral potential, the world’s best geodata, an excellent infrastructure and an operating environment supported by the government’s pro-mining policy. Readily available, high quality geodata combined with the excellent infrastructure make Finland a cost-effective target for mineral exploration companies.
- Apr 2017
Mineral potential targeting using geographical information system is an efficient technique to delimit a study area for further exploration of mineral deposits. This introduction presents an overview of the mineral potential modeling methods and future perspectives of research in the fields of target generation and summarizes the papers that have been incorporated into this Special Issue of the Journal of African Earth Sciences.
- Jan 2017
This paper combines knowledge- and data-driven prospectivity mapping approaches by using the receiver operating characteristics (ROC) spatial statistical technique to optimize the process of rescaling input datasets and the process of data integration when using a fuzzy logic prospectivity mapping method. The methodology is tested in an active mineral exploration terrain within the Paleoproterozoic Peräpohja Belt (PB) in the Northern Fennoscandian Shield, Finland. The PB comprises a greenschist to amphibolite facies, complexly deformed supracrustal sequence of variable quartzites, mafic volcanic rocks and volcaniclastic rocks, carbonate rocks, black shales, mica schists and graywackes. These formations were deposited on Archean basement and 2.44 Ga layered intrusions, during the multiple rifting of the Archean basement (2.44–1.92 Ga). Younger intrusive units in the PB comprise 2.20–2.13 Ga gabbroic sills or dikes and 1.98 Ga A-type granites. Metamorphism and complex deformation of the PB took place during the Svecofennian orogeny (1.9–1.8 Ga) and were followed by intrusions of post-orogenic granitoids (1.81–1.77 Ga). The recent mineral exploration activities have indicated several gold-bearing mineral occurrences within the PB. The Rompas Au-U mineralization is hosted within deformed and metamorphosed calc-silicate veins enclosed within mafic volcanic rocks and contains uranium-bearing zones without gold and very high-grade (>10,000 g/t Au) gold pockets with uraninite and uraninite-pyrobitumen nodules. In the vicinity of the Rompas, a magnesium skarn hosted disseminated-stockwork gold mineralization was also recognized at the Palokas-Rajapalot prospect. The exploration criteria translated into a fuzzy logic prospectivity model included data derived from regional till geochemistry (Fe, Cu, Co, Ni, Au, Te, K), high-resolution airborne geophysics (magnetic field total intensity, electromagnetic, gamma radiation), ground gravity and regional bedrock map (structures). The current exploration licenses and exploration drilling sites for gold were used to validate the knowledge-driven mineral prospectivity model.
Project - Mineral Prospectivity Modeller (MPM)
New ArcSDM - Spatial Data Modeler 5 - code for ArcGIS 10 and ArcGIS Pro platform available for download and testing with sample data sets from GitHub https://github.com/gtkfi/ArcSDM.
How to get started with new ArcSDM python toolbox? Check out our first tutorial video. Got suggestions? Leave a comment and don't forget to subscribe to our arcsdm youtube channel.
Feel free to test the code and please give us feedback!
- Dec 2016
Spatial modelling for prospectivity mapping involves the integration of various geoscientific digital map data. It is essential that the quality of data is high class and that the geological processes involved are well understood. Effects of glacial dynamics and glaciogenic geochemical dispersion need to be taken into consideration when using till geochemistry as one of such input dataset for prospectivity modelling. This paper investigates this issue by developing a fuzzy logic prospectivity model that integrates airborne geophysical data with two different till geochemical datasets. First we use the original geochemical sample set and secondly a spatially corrected dataset that is based on the knowledge of glacial dynamics on the regional scale within the study area. The effect of this correction is tested by comparing the modelling results of both cases using the receiver operating characteristics (ROC) technique to validate the models by using the location of known gold deposits represented by exploration drilling. This study confirms that by taking into consideration the transport caused by the glaciation we can significantly improve the performance of a prospectivity model using till geochemical data.
The Paleoproterozoic Central Lapland Greenstone Belt (CLGB) hosts a number of orogenic gold deposits, the total reported gold endowment being 9.1 Moz. Genetic models of the orogenic gold deposits suggest that the gold source for these deposits is the rocks undergoing metamorphism at depth. If a metamorphic source model is applied to the Kittilä terrane, comprising the core of the Central Lapland Greenstone Belt, potentially as much as 30 times the currently known gold resource was mobilized from the up to 9-km-thick metavolcanic rock sequence comprising the terrane. The size distribution of orogenic gold deposits in gold districts with long exploration histories follows a consistent nonlinear pattern. In comparison, the size distribution of the Central Lapland Greenstone Belt is highly skewed, indicating that the number of gold deposits, including deposits in the 0.5-1.5 Moz size category, remain undiscovered in the belt. The study shows that the Central Lapland Greenstone Belt is underexplored and that the remaining orogenic gold potential of the belt is significant.
A prospectivity model for magmatic Ni–Cu deposits was created by integrating spatially referenced geophysical and geochemical datasets based on a simple and practical exploration model. The study area is the Central Lapland Greenstone Belt, Northern Fennoscandian Shield, Finland. Magmatic nickel deposits are related to rock types that are typically characterized by local magnetic and gravity anomalies. These deposit types can also be a source of nickel, copper and cobalt anomalies in the overlying glacial till cover. This straightforward exploration criterion was translated into a fuzzy logic prospectivity model. The model validation is an essential step in justifying the validity of the prospectivity model. This was accomplished by using receiver operating characteristics (ROC) technique. We used the known Ni–Cu occurrences and deposits as true positive cases and other deposit type locations or random points as true negative cases in the validation process. It appears that the ROC technique provides a robust model validation and optimization technique, providing that suitable validation data exists.
- Oct 2015
In mineral prospectivity mapping (MPM) logistic functions have been widely used to transform mineral exploration data or prospectivity values into the [0, 1] range to generate fuzzified evidential maps or to rank target areas as fuzzy prospectivity models. Recently researchers applied logistic functions to assign fuzzy weights of continuous-value spatial evidence. They assigned fuzzy weights to evidential features without using locations of known mineral occurrences (KMOs) as in data-driven MPM and without discretization of evidential values into some arbitrary classes as in knowledge-driven MPM to overcome exploration bias. However these methods suffer exploration bias resulting from expert judgments in defining slope (s) and inflection point (i) of the logistic function, which are defined by trial and error procedure. In this paper, the application of logistic transformation is demonstrated to assign continuous weights to evidential layers of geochemical and geological data. The weights were assigned without discretization of spatial evidence values and without using the locations of KMOs, while the i and s values of the logistic function were defined by a data-driven way. For this, we applied systems of equations including two equations and two unknown variables (i.e., i and s). Thus by solving the system of equations the two unknown variables, i and s, were defined.
- Feb 2014
- 11. Geokemian Päivät 2014 - 11th Finnish Geochemical Meeting 2014
New mineral exploration methods are developed in the project Novel Technologies for Greenfield Exploration (NovTecEx; funded by Tekes Green Mining Program; 2012-2014). Mineral exploration in regions with thick glacial overburden, peat lands and various conservation areas is very demanding, sensitive and expensive. The aim in this project is to find cost- and eco-effective best practices for mineral techniques and concepts. Research partners of the project are Geological Survey of Finland and the University of Oulu. Essential exploration techniques include concurrent use of geological, geochemical and geophysical surveys. These techniques allow us to locate and thoroughly investigate geological processes responsible for mineral deposits and the indications of ore forming processes. The same techniques can also be applied for direct identification of mineral deposits if the sampling density is high enough. However, the efficiency and success require also high quality data processing and interpretation. Advanced ways to do mineral mapping and exploration are studied and developed in the eight tasks of the NovTecEx project. The tasks cover whole range of exploration procedure in glacial terrain from till sampling to sample processing and analyzing, data processing and modeling, and geophysical field and airborne data interpretation. The arctic areas are sensitive and vulnerable to human activities of any kind. The environmental aspects are considered when selecting the sampling methods, routes in the field and time of the sampling. Low-impact sampling techniques coincident with multiple sample material collection and on-site field analysis will be applied by testing the sampling vehicles and methods. A rugged field computer together with a GPS unit will be used to optimize the route between the sampling sites to increase effectiveness and minimizing impact to the nature. Project will also produce a fast and consistent sample preparation and research method for indicator minerals. The use of methods like self-organizing maps (SOM) and object based recognition of bedrock fractures will be developed in regional scale mineral potential modeling. Geophysical surveys result data from the surface and from greater depths of the bedrock. This data can be used to create a regional geological 3D model, which helps in understanding the geological processes. Application of spatial data mining and spatial analysis techniques on 3D data is new and creates possibilities to compile 3D prospectivity maps. The NovTecEx project generates and develops mineral exploration concepts utilizing methods and best available practices to assess mineral potential or directly locate especially deep seated mineral deposits within northern vulnerable terrains. More effective processing, interpretation and modeling techniques of mineral exploration data enable even larger amounts of data being processed fast. This goal serves especially the companies and organizations conducting grass roots mineral exploration on a regional scale. Techniques are also adoptable into target scale exploration.
The project Ultra Low-impact Exploration Methods in the Subarctic (UltraLIM project; 2013-2015) funded by the Tekes Green Mining Programme compares several ultra light geochemical methods and sampling techniques and aims to find the best practices for sampling and assaying samples from very sensitive subarctic regions. Sample media used in this project are the sediments of the upper parts of soils, organic layers, plants and snow. The research will be conducted on known mineral deposit targets given by Geological Survey of Finland (GTK) and the industry partners of the project. Research partners of the project are GTK and the University of Oulu. There is an increasing demand to develop effective methods for exploration in poorly exposed and deep-seated mineralized bedrock sources under thick sediment cover which are often characterized by complex glacial dispersal patterns. The exploration methods should be applicable particularly in environmentally sensitive subarctic areas and therefore the methods should have a very low impact on the nature. The study is divided into three modules: 1) Selective and/or weak leach geochemical techniques, 2) Biogeochemistry, 3) Snow geochemistry. The sampling and assaying is repeated twice in subsequent years in order to evaluate the repeatability, accuracy and precision of the methods. Also a number of quality control samples will be used. The target areas include variable mineral occurrence types (e.g., Au, base metals, REE, etc.) and variable Quaternary geological conditions (e.g., thick glaciofluvial deposits, moraines, peatlands). Deviations from standard sampling procedure may occur due to geological conditions. The number of sampling points will be about 250 in total. The UltraLIM project will produce information on several low impact geochemical techniques for mineral exploration within glaciated terrains. This research gives reliable and practically tested benchmarks for the use of different geochemical sampling and assay techniques in mineral exploration within the subarctic. The project aims to find the best geochemical practices for locating mineral deposits under thick soil cover. Additional aims include promotion of the use of the methods and results of the project in exploration in Finland and in similar terrains abroad.
- Jan 2014
- 31st Nordic Geological Winter Meeting, January 8-10 2014
Finland is located in the central part of last glaciated area where mineral exploration is challenging due to thick glaciogenic overburden, but also due to large peat land areas and nature reserves. Demand is increasing to develop new applications for regional and target-scale exploration. The Green Mining Programme of the Finnish Funding Agency for Technology and Innovation (Tekes) was launched in 2011 to make Finland a global leader of sustainable mineral industry by 2020 and to increase the number of small and medium size enterprises in the mineral cluster in Finland. The Geological Survey of Finland (GTK) has several ongoing projects within this programme and two of these concern new methodologies for sampling, analysis and interpretation of multiple geological, geochemical and geophysical datasets in environmentally sensitive Arctic and Sub-Arctic areas. The projects are Novel technologies for greenfield exploration (NovTecEx; 2012-2014) and Ultra low-impact exploration methods in the subarctic (UltraLIM; 2013-2015). These projects aim to minimize the environmental impact of mineral exploration, decrease analytical costs, and increase sampling and data interpretation efficiency. In the NovTecEx project new sampling techniques and analytical methods for till geochemistry and indicator minerals are investigated together with advanced data mining methods and interpretation tools for geophysical data. The UltraLIM project focuses on the study and comparison of several geochemical techniques to find the best practices for exploration of various ore types. The three tasks included in the UltraLIM project are: 1) Selective/weak leach techniques, 2) Biogeochemistry and 3) Snow geochemistry.
- Jan 2014
- Geological Survey of Finland
This report presents results of the Central Lapland Greenstone Belt 3D modeling project carried out in 2007–2012. The report documents the utilization of the multiscale edge detection, or “worming”, method for regional gravity data and in prospectivity modeling, presents a new Ni-Cu prospectivity model for the Central Lapland Greenstone Belt (CLGB) district, and documents five 3D geomodels from the Central Lapland area and their implications for the geological interpretation of the area. The worming method was tested on gravity data from Central Lapland. The results indicate that worms, i.e. gravity gradient maxima, are highly useful in interpreting the major shear, thrust, and fault zones, as well as the contacts of lithological units, their depth extent and their dip orientations. Weights of evidence analysis indicates that in the Central Lapland area, the known orogenic gold and iron oxide-Cu-Au deposits display a spatial correlation with the gravity worms, and hence the gravity worms can be used as a tool for locating the most prospective areas for epigenetic mineral deposits. A simple Ni-Cu prospectivity model of the CLGB was constructed using airborne magnetic and ground gravity data together with till geochemistry (Ni, Cu, Co) data. A high-pass filtering technique was used for till geochemistry to filter out regional anomalies and the data were integrated using fuzzy logic. The resulting prospectivity map identifies over 40 target areas for Ni-Cu deposits, including the Sakatti Ni-Cu and Kevitsa Ni-Cu-Au-PGE deposits. A regional 3D model of the Kittilä terrane and the key adjacent structures was constructed using a multidisciplinary approach and a wide array of geophysical and geological data. The modeling results indicate that the Kittilä terrane forms a keel-shaped unit that is ca. 9 km thick at the thickest part, thinning out towards the margins. The shape and thickness suggest that the terrane was considerably thickened during thrusting from the S and NE. A rough estimate of the orogenic gold potential was carried out by applying a metamorphic source model for the gold deposits. Based on the modeling results, potentially up to 228 Moz of gold was mobilized from the Kittilä Group rocks alone during the metamorphic events related to the Svecofennian orogeny. This figure is about 30 times greater than the currently reported gold resources in the known deposits of the area, suggesting that significant undiscovered gold resources remain in the Kittilä terrane area. A 3D model of the Kolari region was constructed using new seismic (HIRE) data from the Hannukainen- Rautuvaara area. The modeling results reveal that the Savukoski and Sodankylä group rocks form a gently SW-plunging open fold structure with an internal small-scale dome and basin structure. The SW–NEstriking Äkäsjoki shear zone has been formed in the axial plane of this fold. The modeling results indicate that the current stratigraphical interpretation is incorrect and needs revising. The results also indicate that the known Fe ± Cu-Au deposits in the area cannot be hosted by the same stratigraphical unit horizon, and are thus not strata-bound deposits, as they have previously been interpreted. A 3D model of the Lapland Granulite Belt (LGB) was constructed using seismic FIRE and geophysical data. According to the model, the LGB consists of at least 4 tectonic blocks. The Vuotso complex immediately SW of the LGB consists of at least 2 tectonic blocks. These units comprise a listric thrust package limiting the Kittilä terrane at its NE contact. Old Outokumpu Oyj drill core and ground geophysical data were used to constrain a deposit-scale model of the Saattopora Au-Cu deposit. The model shows the main shear zones and lithological units, as well as a block model of the two Au-Cu lodes. The modeling data together with the geological observations indicate that the northern ‘A’ lode appears to be controlled by the albitized phyllite unit between a komatiite unit in the south and a mafic tuff unit in the north. The southern ‘B’ lode is controlled by an ESE-striking subvertical shear zone. The modeling data imply that the deposit is open to depth at ca. 160 m below the surface. The data indicate that the ore-hosting veins cross-cut the regional F3 folding visible in outcrops and also in the modeled geological units, suggesting that the mineralization took place in the late stages of the regional deformation and metamorphism. The Lauttaselkä 3D model is based on geophysics and the recent bedrock mapping and drilling campaign in the area. The modeling shows a west-vergent thrust system, with thrust folding explaining the repeating pattern of the Kautoselkä and Vesmajärvi formations in the western part of the area. In the eastern part of the study area, the Salla group rocks on top of the Sodankylä and Savukoski groups is explained by west-vergent thrusting of the Salla group on the latter two units. Keywords (GeoRef Thesaurus, AGI): mineral exploration, gold ores, nickel ores, copper ores, iron ores, gravity methods, edge detection, seismic methods, three-dimensional models, Central Lapland Greenstone Belt, Lapland Granulite Belt, Kittilä, Kolari, Sodankylä, Lapland, Finland
New broadband magnetotelluric (MT) data were acquired in the Central Lapland Greenstone Belt (CLGB) area, northern Finland, during the field campaigns in 2009 and 2010. The measurements belong to an ongoing project at the Geological Survey of Finland. The project intends to create the target- and regional scale 3-D models of the CLGB area using potential field, seismic and electromagnetic data. The survey area is located in the western and northern parts of the CLGB that is one of the largest Proterozoic greenstone belts in the world. In the north and west, the survey area is bordered by Proterozoic granitic rocks and intrusions. The CLGB consists of a Palaeoproterozoic (2.5 - 1.97 Ga) volcanic and sedimentary cover that was deposited on the Archaean (> 2.5 Ga) basement. The Kittilä Group greenstones, which form the core of the CLGB, are suggested representing an allochtonous unit, is bound by tectonic contacts with older units surrounding it (Hanski, 1997). The collected MT dataset consists of the data from 80 sites with the frequency range of 300 - 0.002 Hz and the site spacing of 500 m - 4 km. At the first stage, the MT data were analyzed along a number of crossing 2-D lines. MT parameters were also examined as maps, because the central part of the survey area forms a magnetotelluric array. A regional electrical dimensionality and strike were studied with invariants and various decomposition techniques. Regional electrical dimensionality proved to be mainly 2-D and 3-D except for some northern MT sites in resistive granite-hosted regions, which fulfilled criteria for 1-D interpretation. Smooth 2-D conductivity models were obtained by inverting the determinant of the impedance tensor (Siripunvaraporn & Egbert, 2000; Pedersen & Engels, 2005) and TE- and TM- data jointly using the nonlinear conjugate gradient algorithm of Rodi & Mackie (2001). Model resistivities range from 0.1 Ohm-m to greater than 20 000 Ohm-m in the survey area. The highest conductivities are related to N-S elongated graphite- and sulfide-bearing schists of the CLGB, which are visible also in the airborne electromagnetic data of the study area. Results show that these conductors have the deep roots of about 5 - 10 km. The highest resistivities emerge from granite intrusions that are located in the northern part of the study area. In the west, the resistivity of the CLGB is much higher with no indications of high-conductivity anomalies in the uppermost 40 km. However, in the westernmost part of the study area, a conductivity contrast is observed at the depth of about 10 km possibly indicating the contact zone of the two cratons of the Fennoscandian Shield, i.e. the Karelian and the Norbotten cratons (Lahtinen et al., 2005). 2-D inversion models are presented together with seismic data from the Finnish Reflection Experiment (FIRE) along the CMP-lines 4A and B. In the eastern and central parts of the study area, conductivity anomalies are usually associated with dipping reflectors, whereas such a relationship is not evident in the western part of the study area.
Forty-seven metallogenic areas have been identified in Finland. Of these, 10 areas are dominantly potential for ferrous metals (Fe, Ti, V, Cr), 11 for precious metals (Au, Pd, Pt), 11 for nickel, 8 for copper, zinc and/or lead, 4 for metals mostly used in advanced technologies ('hightech metals' Be, Li, Nb, REE, Ta), and 3 for uranium. Many of the metallogenic areas are potential for more than just one major group of metals. The Finnish metallogenic areas include more than 30 different genetic types of metal deposits. By past production and present resources, the most significant deposit types include: mafic intrusion-hosted Ti-Fe-V (e.g., Mustavaara deposit at Koillismaa), mafic to ultramafic hosted Cr (Kemi), IOCG-style Fe±Cu,Au (Hannukainen deposit in the Pajala-Kolari area), magmatic Ni-Cu-PGE (Portimo, Koillismaa, Hitura and Kotalahti areas, and Kevitsa and Sakatti deposits), orogenic gold (Kittilä, and VMS (Vihanti-Pyhäsalmi). Highly significant are also the unique deposit types of Outokumpu Cu-Co and Talvivaara Ni-Zn-Cu-Co. Most of the known metal endowment of Finland was formed during the Palaeoproterozoic Era, during 2.45-1.92 Ga multi-stage rifting and the 1.9-1.8 Ga Svecofennian orogeny. Detected metal endowment in the Archaean is relatively low with minor komatiite-related Ni (Kuhmo) and orogenic gold deposits (Ilomantsi). Carbonatite-hosted Nb-REE at Sokli, dated to ca. 365 Ma, is the main post-Svecofennian metal deposit known from Finland.
- Jan 2011
This paper demonstrates the use of a fuzzy logic spatial modelling technique in mineral prospectivity mapping. We have used structural data derived from 3D modelling of reflection seismic and potential field surveys to further develop the previously published prospectivity mapping results for orogenic gold deposits within the Central Lapland Greenstone Belt, Northern Fennoscandian Shield, Finland. The fuzzy logic technique in GIS provides a flexible tool to test a conceptual exploration model when there is good data coverage within the area of interest. Furthermore, the results of this exploration model testing can be statistically validated if there are a number of known mineral occurrences of the sought mineral deposit type in the study area. In this study, model validation was conducted using the receiver operating characteristics (ROC) technique
- Jan 2011
- Eleventh Biennial SGA Meeting
Niiranen, Tero; Nykänen, Vesa; Lahti, Ilkka; Karinen, Tuomo 2011. Gravity data in regional scale 3D and gold prospectivity modelling - example from the Central Lapland greenstone belt, northern Finland [Electronic resource]. In: Let's talk ore deposits : proceedings of the Eleventh Biennial SGA Meeting, 26-29th September 2011, Antofagasta, Chile. Antofagasta: Ediciones Universidad Católica del Norte, 265-267. Optical disc (CD-ROM).
Mapping of mineralized geologic structures using geophysical potential field datasets has become an essential part of present-day exploration projects. Various geophysical processing and semiautomatic interpretation techniques have provided new tools into the field of conventional exploration process. Such is the multiscale edge detection or "worming-technique" introduced by Hornby et al., (1999). Worms are representations of the maxima of potential field horizontal gradients. They are calculated at different upward continuation levels providing an alternative view into potential field anomalies and geometry of the anomaly sources. In this work we use the worming-technique on the regional gravity dataset collected by the Geological Survey of Finland during the last four decades in the northern Finland. The dataset consists of more than 19 000 ground gravity observations covering an area of about 15000 km2 with an average site separation of 0.5 - 1 km. The study area covers the central part of the 2.4-2.0 Ga Central Lapland Greenstone belt (CLGB) which is one of the largest Proterozoic greenstone terrains in the world. The CLGB hosts numerous gold occurrences of varying type and size. The majority of the gold occurrences fall into the orogenic gold category but also Iron Oxide-Copper-Gold (IOCG) and paleoplacer types are known within the region (e.g. Eilu et al., 2007). Currently the largest known deposit in the area is the Suurikuusikko orogenic gold deposit with current resources exceeding 5 million ounces Au. The largest known gold resources in IOCG type deposit is in the Hannukainen deposit with ca. 200 000 ounces of gold. All the known orogenic gold and IOCG deposits in the CLGB show intimate spatial correlation to shear zones of varying scale. Processed gravity worms display striking spatial correlation with the known orogenic gold and IOCG deposits. In some cases the gold hosting shear zones are outlined by gravity worms either completely (Sirkka shear zone) or partly (Kiistala shear zone) whereas the presence of some major shear zones is indicated by truncation of worms at the location of the shear zone (Hanhimaa and Muusa shear zone). The important part of our work is the evaluation of the spatial correlation of gravity worms and known gold deposits using the weights-of-evidence calculation procedure. Our results show that presence of gravity worms are indicative of the structures controlling most epigenetic gold deposits in the CLGB area. Therefore, worming-technique proved to be an excellent tool in mapping of structures being prospective for epigenetic gold deposits in the study area and therefore likely in other areas of similar geology. References: Eilu, P., Pankka, H., Keinänen, V., Kortelainen, V., Niiranen, T., Pulkkinen, E., 2007. Characteristics of gold mineralisation in the greenstone belts of northern Finland. In: Gold in the Central Lapland Greenstone Belt. Geological Survey of Finland. Special Paper 44. Espoo: Geological Survey of Finland, 57-106. Hornby, P., Boschetti, F. & Horowitz, F., 1999. Analysis of potential field data in the wavelet domain. Geophysical Journal International, 137, 175 - 196.
- Sep 2008
This thesis comprises a synopsis and six original papers dealing with mineral prospectivity mapping of gold within the Northern Fennoscandian Shield, Finland. Geographical information systems (GIS) with weights-of-evidence, logistic regression, fuzzy logic and neural network models were used in these papers to complete a series of spatial modelling tasks to assess the gold potential of Northern Finland. These results can be used to define regional scale target areas for gold exploration.
- Mar 2008
Among the more popular spatial modeling techniques, artificial neural networks (ANN) are tools that can deal with non-linear relationships, can classify unknown data into categories by using known examples for training, and can deal with uncertainty; characteristics that provide new possibilities for data exploration. Radial basis functional link nets (RBFLN), a form of ANN, are applied to generate a series of prospectivity maps for orogenic gold deposits within the Paleoproterozoic Central Lapland Greenstone Belt, Northern Fennoscandian Shield, Finland, which is considered highly prospective yet clearly under explored. The supervised RBFLN performs better than previously applied statistical weights-of-evidence or conceptual fuzzy logic methods, and equal to logistic regression method, when applied to the same geophysical and geochemical data layers that are proxies for conceptual geological controls. By weighting the training feature vectors in terms of the size of the gold deposits, the classification of the neural network results provides an improved prediction of the distribution of the more important deposits/occurrences. Thus, ANN, more specifically RBFLN, potentially provide a better tool to other methodologies in the development of prospectivity maps for mineral deposits, hence aiding conceptual exploration.
Despite its exploration immaturity, the Central Lapland Greenstone Belt has geodynamic, geological and gold-endowment characteristics to indicate that it is potentially the premier orogenic-gold province in Finland. Previous GIS-based prospectivity mapping of the Central Lapland Greenstone Belt used only empirical geophysical and geochemical datasets based on the weights-of-evidence integration method. In this study, such empirical data are combined with conceptual geological parameters derived from a 1:200 000 scale bedrock map of northern Finland and use the weights-of-evidence, logistic regression and fuzzy-logic methods for integration. A targeting model is developed based both on the empirical data and conceptual parameters defined from understanding of the processes essential to produce large deposits in an orogenic-gold mineral system. Key spatially referenced layers are derived from the empirical data and from critical proxies for essential processes in the targeting model, and the spatial association between these key layers and a training set of known gold deposits and prospects is quantified. Key parameters are then reclassified into binary layers according to maximum spatial association with the training set, and these layers are then integrated into prospectivity maps using both the weights-of-evidence and logistic regression methods. The highest probability class in all models defines
- Feb 2008
The conceptual approach used in this study incorporates spatial analysis techniques for data integration and analysis to perform reconnaissance-scale mineral prospectivity mapping for iron oxide copper – gold (IOCG) mineralisation in Finland. The known IOCG occurrences in Finland are characterised by the following features: (i) an epigenetic magnetite-rich host-rock; (ii) an association of Fe – Cu – Au ± Co ± U; (iii) ore minerals comprising magnetite, chalcopyrite, pyrite or pyrrhotite, and native gold; (iv) a gangue dominated by Ca-amphibole ± diopside, albite and biotite; (v) enrichment in Ag, Au, Bi, Ca, CO2, Cu, Fe, S, Te ± As, Ba, Cl, Co, K, LREE, Mo, Na, Pb, Rb, Sb, Se, U; (vi) multi-stage alteration; (vii) formation in the P – T range of 400 – 600°C, 150 – 350 MPa; and (viii) a distinct structural control in regions that have experienced both extensive compression and extension. The datasets used for the prospectivity analysis include a 1:1 000 000 scale geological map, high-resolution airborne geophysics, regional-scale multi-element till-geochemistry data, and a mineral occurrence database. The derived parameters used in the conceptual analysis include: (i) proximity to the craton margin; (ii) intersecting fault structures; (iii) presence of granitic intrusions particularly those with compatible and incompatible element enrichment; (iv) Cu, Co and Fe concentrations in till samples; (v) presence of hematite; and (vi) airborne magnetic highs and radiometric U data. A conceptual fuzzy-logic model was used to predict and locate the most prospective or favourable areas for IOCG exploration in the study area using the above-mentioned data layers. The models identify several permissive and high-potential areas within a significantly reduced potential exploration area. Validation of the modelling was conducted by quantifying the spatial association between the predicted endowment as favourability classes on the prospectivity map and the known mineral deposit sites with IOCG affinities using the Bayesian weights-of-evidence method.
- Jan 2008
Gold prospectivity of Finland was assessed to refine exploration criteria for future work on different deposit classes or styles. For the prospectivity assessment, Finland was divided into geologically different terrains which are SW Finland (including Tampere Schist Belt and Vammala Migmatite Zone), Ilomantsi (Hattu Belt), Pohjanmaa, Kuhmo and Suomussalmi greenstone belts, Kuusamo Schist Belt, Peräpohja Schist Belt, and Central Lapland. The specific modelling criteria suggested for each type of mineralisation are summarised at the end of the report. The general criteria and comments regardless of mineralisation type include: 1) A suitable seal/trap, a suitable fluid conduit exists, and a suitable element source should be present. 2) Regional contacts between volcanic and sedimentary rocks should be included in all models. 3) Structural data should be incorporated into most data sets. 4) Geochemical gradients should be used in preference to actual anomaly values. 5) Genetic classification of the deposits is important as empirical GIS models should only use similar deposits as learning points. 6) Schist areas can probably be excluded from modelling, with emphasis instead placed on sediment/volcanic belts and contacts. 7) Dating of deposits is required to compare mineralisation ages to the age ranges of crustal formation events. 8) Defining tectonic settings helps to define the likely deposit types to exist. 9) Areas with thin lithosphere and major boundaries are important. 10) Defining the distribution of anomalous intrusion types to help identify tectonic environments and crustal depths exposed.
Palaeo-stress modelling applies geomechanical modelling codes to define dilatant (low stress) sites in a rock mass. These dilatational sites can be considered targets for structurally controlled mineralization. The main assumptions in the conceptual model are that the strain pattern of faults and lithological contacts has not been significantly changed since mineralization ie. that mineralization is late in the structural history, and that stress orientations and magnitudes and rheological properties of faults and rock units are known. UDEC code was applied to determine the palaeo-stress pattern caused by the fault network geometry and rheology within the Central Lapland Greenstone Belt at 1:250 000 scale to target orogenic gold mineralization. Modelling simulated E-W, N-S, NE-SW and NW-SE compression directions (D3-D4 deformation stage) and the minimum principal stress (σ3) is used to indicate dilational areas. The spatial association between the palaeo-stress modelling results of 12 different models (compression orientations and different fault classifications) and the known gold occurrences was quantified using logistic regression methods in ArcGIS. The minimum principal stress distributions of the models, which simulate E-W compression, have the highest positive weight, contrast and confidence. The maximum spatial association pattern in the best model cover 1.3% of the study area and include 11% of the known gold occurrences. In addition, the models that simulate NE-SW compression simulations give a good spatial association with known gold occurrences. This is in a good agreement with structural studies of the Central Lapland Greenstone Belt, which suggest that far field stress has varied from NE-SW to E-W during the late stages of the tectonic evolution and orogenic gold mineralization.
The Geological Survey of Finland has done a lot of work to explore gold in the Central Lapland Greenstone Belt in northern Finland. The whole area is proven to have potential for Au mineralization, and of that the two presently active gold mines are the proof. New methods have been used to define the most prospective areas for gold exploration in this district. GIS-based prospectivity analysis is one of them. Four target areas (Vuomanperänmaa, Nuttiot, Petäjäselkä and Lauttaselkä) anomalous in analysis have been selected for detailed studies. An aim has been to use surficial geological methods to examine the result of analysis and also focus future exploration. Test pit studies with till geochemical and heavy mineral sampling proved to be effective and fast methods to get information of the till deposits and for estimating transport distance of mineralized material under former glacial processes. The results of this study strengthen the decision to continue gold exploration in all of the targets.
Geoscientific Information Systems (GIS) provide tools to quantitatively analyze and integrate spatially referenced information from geological, geophysical, and geochemical surveys for decision-making processes. Excellent coverage of well-documented, precise and good quality data enables testing of variable exploration models in an efficient and cost effective way with GIS tools. Digital geoscientific data from the Geological Survey of Finland (GTK) are being used widely as spatial evidence in exploration targeting, that is ranking areas based on their exploration importance. In the last few years, spatial analysis techniques including weights-of-evidence, logistic regression, and fuzzy logic, have been increasingly used in GTK’s mineral exploration and geological mapping projects. Special emphasis has been put into the exploration for gold because of the excellent data coverage within the prospective volcanic belts and because of the increased activity in gold exploration in Finland during recent years. In this paper, we describe some successful case histories of using the weights-of-evidence method for the Au-potential mapping. These projects have shown that, by using spatial modeling techniques, exploration targets can be generated by quantitatively analyzing extensive amounts of data from various sources and to rank these target areas based on their exploration potential.
Weights-of-Evidence (WofE) and Radial Basis Function Link Net (RBFLN) were applied to soil group mapping in eastern Finland. The data consisted of low altitude airborne geophysical measurements, Landsat 5 TM-satellite image, and digital elevation model (DEM) and slope information derived from it. Probability maps were constructed for each soil group one by one and combined into a prediction map of soil groups using maximum posterior probability (WofE) or pattern membership (RBFLN). Self-Organizing Map (SOM) and Sammon’s Mapping were applied for selecting the data sets for modeling and visualizing the data. The soil types belonging to each soil group used in the Arc-SDM modeling were defined by clusters revealed by the SOM and Sammon’s Mapping algorithms. The soil types with similar characters were collected in the same cluster. Numerical evaluation of the models’ performance was performed using the confusion matrix. The Ratio of Correct Classifications (RCC) for the best WofE model was 0.64 in the training area and 0.61 in the testing area. The RCC for the best RBFLN model was 0.62. Modeling of soil groups using Arc-SDM is time consuming because models need to be constructed for each soil group before combining them into a final prediction map. In this study a simple method was tested for combining the maps. In the future, more attention should be paid to combining the posterior probability models and also to selecting data sets used for modeling.
nykänen,v. &salmirinne, h 2007. Prospectivity analysis of gold using regional geo-physical and geochemical data from the Central Lapland Greenstone Belt, Finland Geological Survey of Finland, Special Paper 44, 251–269, 9 figures and 8 tables. Two spatial modeling techniques, empirical weights of evidence and conceptual fuzzy logic, were used to predict the most prospective areas for gold exploration within the Paleoproterozoic Central Lapland Greenstone Belt in Northern Finland. The study area covers almost 20 000 km 2 in a terrain with excellent infrastructure and easy access in spite of situating in a region some 100 km north from the Arctic Circle. For the empirical model the spatial association between the known mineral occurrences and selected evidential geoscientific datasets were quantified and used to generalize the original datasets into binary predictor patterns indicating favorable areas. The conceptual model, on the other hand, was created by using expert opinions on the significance of the anomalies within the evidential datasets. The data used were high-resolution airborne geophysics, regional gravity, and regional scale multi-element till geochemistry. Both methods, empirical and conceptual, give comparable results, which were geologically validated and verified to be meaningful. The models used predicted well the known deposits and highlighted also areas with past and future exploration interest. Both empirical and conceptual approaches proved to be suitable for the purpose and the regional geochemical, geophysical and geological data appeared to be appropriate for regional or reconnaissance scale first stage exploration. According to the evidential data used and the models produced, the Central Lapland Greenstone Belt has regions with high potential for new greenstone hosted orogenic type of gold occurrences to be found.
- Dec 2006
A recently published study has shown that small-scale geologic map data can reproduce mineral assessments made with considerably larger scale data. This result contradicts conventional wisdom about the importance of scale in mineral exploration, at least for regional studies. In order to formally investigate aspects of scale, a weights-of-evidence analysis using known gold occurrences and deposits in the Central Lapland Greenstone Belt of Finland as training sites provided a test of the predictive power of the aeromagnetic data. These orogenic-mesothermal-type gold occurrences and deposits have strong lithologic and structural controls associated with long (up to several kilometers), narrow (up to hundreds of meters) hydrothermal alteration zones with associated magnetic lows. The aeromagnetic data were processed using conventional geophysical methods of successive upward continuation simulating terrane clearance or ‘flight height’ from the original 30 m to an artificial 2000 m. The analyses show, as expected, that the predictive power of aeromagnetic data, as measured by the weights-of-evidence contrast, decreases with increasing flight height. Interestingly, the Moran autocorrelation of aeromagnetic data representing differing flight height, that is spatial scales, decreases with decreasing resolution of source data. The Moran autocorrelation coefficient scems to be another measure of the quality of the aeromagnetic data for predicting exploration targets.
- Aug 1995
Chromites from the ultramafic rocks of the early Proterozoic Outokumpu-Jormua Ophiolite Belt were found to possess textural and and ratios that resemble those described in Mesozoic ophiolite complexes. Chromites of cumulus texture were found at all locations along the Belt, whereas residual and podiform chromites were found only in the Outokumpu complex. The chromites in the Outokumpu complex originated in a supra-subduction zone environment, whereas those in the Jormua ophiolite complex originated in a mid-ocean ridge environment. The main difference between the early Proterozoic ophiolitic chromites and those of Mesozoic age is the relatively high Zn content in the former, which is a metasomatic feature related to the presence of large amounts of Cu-Co-Zn-Au ore. The data and observations presented provide evidence contributing to the concept of Proterozoic plate tectonics.
- Apr 1995
Chromite deposits are classified into stratiform and podiform types according to their shapes and textures. Podiform deposits are hosted by alpine-type peridotite, contrasting with the regularly stratefied chromite layers of the stratiform complexes. This paper reports the first discovery of small but scientifically interesting podiform chromite occurence in the Outokumpu Ophiolite Complex in eastern Finland. The occurrence, mineralogy, mineral chemistry, and platinum-group element (PGE) distribution testify to the podiform nature of these chromites and their geodynamic environment.
- Nov 1994
The Palaeoproterozoic sedimentary cover of the Archaean basement complex in the Fennoscandian Shield is intruded by several stages of mantle-derived magmas. The 2.1 Ga Fe-tholeiitic magmatism, which is characterized by hypabyssal intrusions, dykes and volcanic flows intersecting the cover, also forms a widespread dyke swarm within the basement in North Karelia, eastern Finland. There are certain differences between these age-related sets of dykes, suggesting lithospheric thinning resulting from extensional tectonics during 2.1 Ga events.REE abundances of the Fe-tholeiitic dykes within the basement range from 10 to 40 times chondritic, with gently dipping REE patterns [La/Yb]N=1.2 to 3.8). The dykes intersecting the epicontinental sediments of the cover have total REE abundances ranging from 10 to 70 times chondritic, with LREE-enriched, gently or steeply dipping REE patterns ([La/Yb]N=2.8 to 4.1), whereas the dykes intersecting the shallow-marine sediments within the cover have total REE abundances from 5 to 55 times chondritic, with almost flat REE patterns ([La/Yb]N=1.2 to 1.8).On the basis of the REE patterns and modelling the melting it is suggested that the melt generation occurred in a transitional tectonic setting, where both garnet- and spinel-bearing mantle lherzolites were involved due to lithospheric thinning during Palaeoproterozoic extensional tectonism.
Investigations into chromite compositions in Archaean ultramafic rocks are rare. Most Archaean ultramafic rocks have undergone a multi-stage history of deformation and metamorphism, and therefore fresh primary chromite is rarely preserved. The preserved cores nevertheless offer a method for studying the petrogenesis of highly altered Archaean ultramafic rocks. Hundreds of electron microprobe analyses have been performed on chrome spinels in the serpentinites from the Archaean Kuhmo greenstone belt to constrain the origin of the serpentinite massives. Best preserved chromites can be divided into two groups on the basis of their composition. Those of the first group share the characteristics of Alpine-type complexes in terms of Cr/(Cr+Al) values from 0.58 to 0.76 and Mg/(Mg+Fe2+) values from 0.53 to 0.60, while those of the second type have altered to ferritchromite and have their Mg/(Mg+Fe2+) values from 0.00 to 0.37. These observations suggest that lithological unit composed of serpentinites, and pillow lavas of tholeiitic (Thl), komatiitic, and basaltic komatiitic composition, known as the Kellojärvi group, includes parts of an Archaean ophiolite.
A description is given of Mn-Mg-rich members of the ilmenite series from the serpentinized wehrlite from Ensila, Kuhmo greenstone belt, Finland. Microprobe analyses give a composition of Il17He1Py53Ge29 for the Mn-Mg-richest ilmenite. -Authors