Knowledge Recovery for Continental-Scale Mineral Exploration by Neural Networks

This study is concerned with understanding of the formation of ore deposits (precious and base metals) and contributes to the exploration and discovery of new occurrences using artificial neural networks. From the different digital data sets available in BRGM's GIS Andes (a comprehensive metallogenic continental-scale Geographic Information System) 25 attributes are identified as known factors or potential factors controlling the formation of gold deposits in the Andes Cordillera. Various multilayer perceptrons were applied to discriminate possible ore deposits from barren sites. Subsequently, because artificial neural networks can be used to construct a revised model for knowledge extraction, the optimal brain damage algorithm by LeCun was applied to order the 25 attributes by their relevance to the classification. The approach demonstrates how neural networks can be used efficiently in a practical problem of mineral exploration, where general domain knowledge alone is insufficient to satisfactorily model the potential controls on deposit formation using the available information in continent-scale information systems.     

Internal structure of Piton de la Fournaise volcano from seismic wave propagation and earthquake distribution

Arrival times of seismic waves from local earthquakes are inverted for both locating the source and defining the 3-D velocity heterogeneity of Piton de la Fournaise.The lateral heterogeneity of the 2632 m high edifice is resolved as a high-velocity plug, 1.5 km in diameter, surrounded by a low-velocity ring, which may be interpreted as due to the construction of Fournaise on the flank of the older volcano Piton des Neiges. Wave mode conversion detected on three-component seismograms provides evidence for boundaries of contrasted velocities.Pre-eruptive swarm earthquakes cluster in the high-velocity zone, under the Dolomieu summit crater. Low strength and cohesion of the surrounding material account for the lack of seismicity for the final 1–3 km radial flow of magma to the vents in Enclos Fouqué.Beneath the high-velocity plug the existence of a body with low velocity for P, and even for S, waves is well constrained. However, the walls and base are poorly defined because of the lack of deep earthquakes for sampling. The few earthquakes that are located in this depth region usually occur at a depth of around 1.5 km below sea level in the region of the cone. This can be considered providing the upper constraint on the lower limit of the aseismic part of the low-velocity body. The coincidence in time of their occurrence with the swarms above sea level and the eruptions suggests magmatic activation of the low-velocity aseismic volume 1.5 km below sea level under the high-velocity plug of the cone. Further down, the concentration of seismicity in two swarms, between 2 and 4 km, under the eastern flank does not allow the structure to be sampled effectively.     

Subgrain-scale decoupling of Re and Os and assessment of laser ablation ICP-MS spot dating in molybdenite

Reproducibility of Re-Os molybdenite ages depends on sample size and homogeneity, suggesting that Re and Os are decoupled within individual molybdenite crystals and do not remain spatially linked over time. In order to investigate the Re-Os systematics of molybdenite at the subgrain (micron) scale, we report LA-ICP-MS Re-Os ages for an Archean molybdenite crystal from Aittojärvi, Finland, analyzed in situ in a white aplite matrix. A related Aittojärvi molybdenite (A996D), in the form of a very fine-grained mineral separate, is used as one of our in-house NTIMS standards, and thus its age of 2760 ± 9 Ma is well established. Measurements of (¹⁸⁷Re + ¹⁸⁷Os)/¹⁸⁵Re on micron scale spots along 200 μm traverses across the crystal yield a wide range of ages demonstrating that, in this case, microsampling of molybdenite does not produce geologically meaningful ages. Experimentation with mineral separations and sample size over a 7-yr period predicted that this would be the outcome. We suggest that ¹⁸⁷Os is more likely to be the mobile species, based on its charge and ionic radius, and that ¹⁸⁷Os becomes decoupled from parent ¹⁸⁷Re with time on the micron and larger scale. Incompatible charge and ionic radius for Os ions formed during reduction of molybdenite-forming fluids may explain the widely observed absence of common (initial) Os in molybdenite. Geologically accurate ages for molybdenite can only be obtained for fully homogenized crystals (or crystal aggregates) so that any post-crystallization ¹⁸⁷Re-¹⁸⁷Os decoupling is overcome.A growing number of geologically accurate ID-NTIMS ¹⁸⁷Re-¹⁸⁷Os ages for homogenized molybdenite suggest that postcrystallization mobility of radiogenic ¹⁸⁷Os must be limited to within the molybdenite mineral phase. We suggest that radiogenic ¹⁸⁷Os may be stored in micron scale dislocations, kink bands, and delamination cracks produced by deformation, and that the unusual structure and deformation response of molybdenite results in an increased chemical stability in this mineral. Migration of ¹⁸⁷Os into adjacent silicate phases is highly unlikely, but other contacting sulfides may take in Os. In an example from a Proterozoic skarn deposit at Pitkäranta (western Russia), we demonstrate minor loss of radiogenic ¹⁸⁷Os from molybdenite and a corresponding gain in adjacent chalcopyrite such that the molybdenite age is not perceptibly disturbed, whereas the resulting chalcopyrite ages are impossibly old. Therefore, it is unadvisable to perform Re-Os analytical work on any sulfide in contact or intimate association with molybdenite. In addition to large errors in the age, if the isochron method is employed, initial ¹⁸⁷Os/¹⁸⁸Os ratios could be erroneously high, leading to seriously errant genetic interpretations.     

Sur les changements du gradient vertical de temperature dans le brouillard de Rayonnement

Résumé Des sondages aérologiques, par ballon captif équipé d'un météorographe, au sein d'une couche de brouillard de rayonnement dense et particulièrement persistante (épaisseur 300 m, durée 11 jours) ont permis à l'auteur de mettre en évidence d'importants changements internes du gradient de température. Il les décompose en trois phases d'évolution successives suivant la densité du brouillard: 1) inversion; 2) isothermie et 3) gradient approximatif de l'adiabatique saturée. De plus, il met en évidence l'existence éphémère de tranches d'air à gradient thermique nettement suradiabatique se formant au sommet de la couche nuageuse par suite du rayonnement et de l'évaporation vers l'espace. On émet l'hypothèse que ces tranches instables sont mobiles vers le bas et servent à refroidir l'ensemble de la couche de brouillard d'où, par la suite, changement progressif du gradient thermique interne.

---

Summary During a long period of dense radiation fog (11 days, 1000 ft. thick), a meteorograph attached to a captive balloon was used to investigate the vertical temperature-lapse-rate. The records reveal three consecutive phases in the development of the fog: i. inversion; ii. isothermal state; iii. approximately saturated-adiabatic lapserate.Owing to radiation and evaporation, the top layer often shows instability. The results of the observations seem to support the author's hypothesis that the instable layer gradually spreads downward while the temperature decreases in the whole fog.

---

Zusammenfassung Während einer besonders lang andauernden Nebelperiode (11 Tage) waren in der dichten Nebelschicht (300 m) aerologische Sondierungen mit Hilfe eines an einem Fesselballon befestigten Meteorographen durchgeführt worden, die erlaubt hatten, grundlegende Veränderungen des Temperaturgradienten festzustellen. Je nach Dichte des Nebels können 3 Entwicklungsstufen beobachtet werden: 1. Inversion; 2. Isothermie und 3. beinahe Feuchtadiabate. Ueberadiabatische Gradienten, die sich an der Nebeloberfläche durch Ausstrahlung und Verdunstung gegen die freie Atmosphäre bilden, werden besonders hervorgehoben. Es wird vermutet, dass sich diese instabilen Luftpakete nach unten bewegen, die ganze Nebelschicht auskühlen und somit eine progressive interne Aenderung des Temperaturgradienten hervorrufen.

     

The sulfur isotope composition of basaltic rocks

The sulfur isotope composition of tholeiitic basalts, olivine alkali basalts and alkalirich undersaturated basalts were investigated. A method of preparation was devised

1. for the extraction of the small amounts of sulfur contained in the rock samples (about 100 ppm S),
2. for the separation of sulfide- and sulfate-sulfur.

Tholeiitic and olivine alkali basalts show a predominance of sulfide-sulfur. Alkali-rich undersaturated basalts show sulfide- and sulfate-sulfur. The oxidation potential of the magma is reflected in the proportions of sulfide- and sulfate-sulfur. Differences in the conditions of oxidation are also the cause of the sulfur isotope fractionation observed. The mean in the isotope composition of the sulfur in the olivine alkali basalts (with the exception of two samples which show extreme deviation) is δ ³⁴S= +1.3 per mil. The values for the olivine alkali basalts are concentrated around this mean in a remarkable way, showing only small deviation for the individual samples. When the tholeiitic basalts deviate from this mean, it is only with a relative enrichment in the ³²S isotope. With a pronounced variation of the individual values, the mean for the sulfide-sulfur is δ ³⁴S=?0.3 per mil. The few sulfate values of both types of basalt are without significance for the discussion of their origin. However, this does not apply to the alkali-rich undersaturated basalts. Due to the higher water content, this basaltic magma had a higher oxygen partial pressure which favoured the formation of SO₂ and SO ₄ ^2? besides H₂S while pressure was released during the ascent of the magma. The sulfur isotope fractionation connected with this oxidation led to a total enrichment of ³⁴S in the rock, (δ ³⁴S for total sulfur: +3.1 per mil) with particular favouring the sulfate (δ ³⁴S=+4.2 per mil). It is accepted that the sulfur of all three types of basalts derives directly from the mantle. The olivine alkali basalts show the least deviation from the mantle value, which, in the place of origin of the basalts from the region investigated, would probably have been δ ³⁴S=+1.3(±0.5) per mil. From this it may be concluded that the olivine alkali basalts — the most frequent type of basalt in this region — had their origin in the partial melting of the mantle without further differentiation. From the sulfur isotope data we concluded that the primary isotope composition of the continental tholeiitic basalts probably corresponds to that of the olivine-alkali basalts, and to that of the mantle. However, due to degasing in the layers near to the surface, some samples lost ³⁴S, which may be related to the formation of SO₂ during the release of pressure. There is no positive indication of a differentiation in shallow depths (<15 km — in the sense of Green and Ringwood, 1967). The reason for the obvious isotopic fractionation of the alkali-rich undersaturated basalts may be seen in their higher primary water content. This is a pronounced indication of the origin of this type of magma. Bultitude and Green (1968) proved by experiment, that the formation of alkali-rich undersaturated basaltic magma is possible in the mantle in the presence of water. Only a small amount of water is available for the formation of magma in the mantle. With a water content higher than normal for basalts, only small amounts of magma can be formed, but at lower temperatures this would allow the melting of a larger fraction of mantle material. By reaction with the wall rock, these magmas could be enriched in those components of mantle minerals which have the lowest melting point. This may help to explain their geochemical characteristics.     

Late-Precambrian correlation and the relationship between the Damara and Nama Systems of South West Africa

Recent findings about the late-Precambrian stratigraphic succession in the areas between the Orange River and the Klinghardt Mountains and south of Windhoek throw new light on the relationship between the Damara and Nama Systems.In the Orange River area the formations of the Gariep Group are unconformably or pseudo-conformably overlain by the Nama System. In central and northern South West Africa a similar relationship is exhibited by the Nosib Group and overlying Damara System.Pronounced similarities between the sedimentation and deformational history of both Nosib and Gariep indicate a probable correlation of the two groups and this result is also supported by radiometric ages.It follows that Damara and Nama are most likely either completely or partly coeval. The latter may represent a shelf facies of the former geosynclinal succession.The Gariep and Nosib deformation apparently started with the subsidence of their respective troughs and may have lasted from approximately 850 m. y. to 600 m. y. This tectonic event is older than the Damara Orogeny (550 to 450 m. y.) and may be time-equivalent to the Katanga Episode.Two late-Precambrian glaciations were widespread in South West Africa and their sedimentary deposits are important for the correlation of Gariep/Nama with Nosib/Damara.

---

Zusammenfassung Neue Ergebnisse über die spätpräkambrische Gesteinsabfolge im Gebiet des Orange-Flusses und im Rehoboth-Distrikt südlich von Windhoek sind von großer Bedeutung für das Altersverhältnis zwischen Damara- und Nama-System.Im Orange-Gebiet werden die Formationen der Gariep-Gruppe diskordant oder pseudo-konkordant vom Nama-System überlagert, während in Zentral- und Nord-Südwest-Afrika das Damara-System ebenfalls diskordant oder pseudokonkordant auf die Nosib-Gruppe folgt. Sedimentations- und deformationsgeschichtliche Ähnlichkeiten deuten auf eine Korrelation zwischen Nosib und Gariep hin, und diese Interpretation wird durch radiometrische Altersbestimmungen unterstützt. Daraus folgt, daß Damara und Nama wahrscheinlich einander entsprechen, wobei die Nama-Sedimente als Plattform-Fazies der geosynklinalen Damara-Serien anzusehen sind.Die Gariep-Deformation begann schon mit dem Einsinken des Gariep-Ablagerungstroges und dürfte von ca 850-600 M. J. angedauert haben, wobei es stellenweise zur Granitisation geeigneter Sedimente und Laven kam. Diese Orogenese ist älter als die von ca. 550-450 M. J. dauernde Damara-Phase und scheint zeitlich der auch im Kaokoveld nachgewiesenen Katanga-Episode zu entsprechen.Zwei spätpräkambrische Vereisungsperioden sind weit verbreitet in Südwest-Afrika, und ihre Glazialablagerungen sind ein wichtiges Hilfsmittel für die Korrelation der Gariep-Nama-Abfolge mit der Nosib-Damara-Stratigraphie.

---

Résumé De nouveaux résultats concernant la série de roches du Précambrien récent de la région de l'Orange et du district de Rehoboth au sud de Windhoek se sont révélés de haute importance pour l'étude de relation d'âge entre les systèmes Damara et Nama.Dans la région de l'Orange, le système Nama repose de façon discordante ou pseudo-concordante sur les formations du groupe Gariep; de même, en Afrique Sud-Ouest, centrale et le système Damara prend une position analogue par rapport au groupe Nosib. Des analogies dans l'histoire de la sédimentation et de la déformation permettent de réaliser une corrélation entre le Nosib et le Garieb; des déterminations de l'âge par méthodes radio-électriques appuient cette interprétation. Par conséquent, le Damara et la Nama sont vraisemblablement en rapport de conformité et les sédiments du Nama sont à considérer comme les faciès des séries géosynclinales du Damara.La déformation du groupe Gariep commença déjà avec l'affaissement des auges de dépôt et a probablement duré de 850-600 m. a.; à cette occasion se produisit, par endroit, la granitisation des sédiments et des laves appropriés. Cette orogénèse est plus ancienne que la phase du Damara qui dura de 550-450 m. a. et semble correspondre dans le temps à l'épisode Katanga également dans le Kaokoveld.Deux périodes de glaciation du précambrien récent sont très largement répandues en Afrique du Sud-Ouest et leurs dépôts glacières sont des moyens très importants pour la corrélation de la série Gariep-Nama avec la stratigraphie du Nosib-Damara.

---

Damara Nama Rehoboth Windhoek. Gariep , - , , - Damara , Nosib. Nosib Gariep . , Damara Nama, , , Nama , Damara. Gariep Gariep , , 850 600 . . Damara, 550-450 , , , Katanga, Kaokoveld. - Gariep-Nama — Nosib-Damara.

---

---

Expanded and slightly modified English version of a paper read for the author by Prof. H.Martin at the 61st Annual Congress of the Geologische Vereinigung and entitled: Spätpräkambrische Orogenesen und das Verhältnis zwischen Damara und Nama System in Südwest Afrika.     

Der Schwefelgehalt in Knochen und Zähnen rezenter und fossiler Organismen

The sulfur content of teeth and bones has been determined by a microprobe. Recent samples only contain about 0.05 weight percent of sulfur, independent of a marine or terrestrial origin of the organisms. The total sulfur of these samples is partly related to the organic matter and partly to SO ₄ ^2– -ions replacing PO ₄ ^3– -ions in the crystalline matter of the skeletons. In the living organisms this matter consists of octacalciumphosphatecarbonate (OCPC) (Hayek, 1967), which is transformed only to a small degree into apatite. During diagenesis the process of apatite-crystallisation proceeds, and more sulfate enters the lattice of the phosphate, which leads to an enrichment of sulfur up to about one percent by weight. There is no relation observed between the concentration of sulfur and the age of the organisms. Influences of the environment of diagenesis on different minor compounds entering the phosphates of bones and teeth in different state of composition have been evaluated. As a result the possibility of dating fossils by minor element analysis (F, Mn, SO ₄ ^2– etc.) must be questioned. Differences between freshwater and marine invironment of deposition cannot be observed in the total sulfur of diagenetic phosphates.

---

---

Ich danke Herrn Prof. Dr. K. H. Wedepohl für sein Interesse an der Durchführung dieser Arbeit und für seine aktive Unterstützung. Für die Überlassung von Proben danke ich ganz besonders Herrn Prof. Dr. O. H. Walliser, Herrn Dr. Frank Westphal, Mrs. L. S. StevenSon, Herrn Prof. Dr. H. W. Flügel, Herrn Prof. Dr. W. G. Kühne und Herrn Prof. Dr. K. Krömmelbein; für klärende Diskussionen Herrn Prof. Dr. O. H. Walliser, Herrn Prof. Dr. Dr. h. c. C. W. Correns, Herrn Prof. Dr. K. H. Wedepohl, Herrn Priv.-Doz. Dr. med. Kämmerer und Herrn Priv.-Doz. Dr. M. LieflÄnder; für die Übermittlung von Literatur und Hinweisen Herrn Priv.-Doz. Dr. H. Newesely; und für ihre Unterstützung und Diskussionsbereitschaft allen Mitarbeitern des Geochemischen Institutes der Universität Göttingen.     

Zur Genese des elementaren Schwefels im Gips von Weenzen (Hils)

Zusammenfassung In der Hilsmulde tritt bei Weenzen ein Salzstock durch Kreide- und Tertiärsedimente zutage. Dieser Gips führt an einigen Stellen elementaren Schwefel. Durch Isotopenanalysen wurde festgestellt, daß der elementare Schwefel gegenüber dem Ausgangssulfat an³⁴S verarmt und der Schwefelm Restsulfat entsprechend angereichert wurde. Schwefelfreie Bereiche des Gipses zeigen den für Zechsteinevaporite üblichen ³⁴S-Wert. Der Isotopenunterschied von durch-schnittlich 26% deutet auf bakteriellen Ursprung des Schwefels, weil eine anorganische Reduktion des Sulfates unter den vorauszusetzenden Temperaturen nicht möglich ist. Es wird angenommen, daß während des Aufstieges des Salzstockes ein geringfügiger Zutritt von Erdöl Lebensbedingungen für anaerobe Bakterien schuf, was zur Reduktion des Sulfats zu H₂S führte. Der dabei freiwerdende Schwefelwasserstoff wurde durch Reaktion mit SO ₄ ^2– in elementaren Schwefel umgewandelt. Da sich die Reduktion in den einzelnen Bereichen des Gipses sicher in Abhängigkeit von der Intensität der Bakterientätigkeit vollzog, hat sie primär wohl nicht zu homogener Verteilung der Schwefelisotope im Schwefelwasserstoff bzw. im Restsulfat geführt. Während für den elementaren Schwefel die nachträgliche Homogenisierung wegen der Beweglichkeit des Schwefelwasserstoffes verständlich ist, kann sie im Restsulfat so erklärt werden, daß dieses während der Neokomtransgression von oben her stärker der Lösung ausgesetzt war und umkristallisierte, wodurch sich die anfangs wahrscheinlich stärkeren Unterschiede in der Isotopenzusammensetzung des Gipses verwischten.

---

Native sulphur is encountered occasionally within the gypsiferous cap-rock of a Permian salt dome in the Hils syncline, in the environs of Weenzen (see map, Fig. 1). A number of sulphur isotope analysis demonstrated that the native sulphur originated out of the primary sulphates, had been depleted in³⁴S, causing a relative enrichment of the remaining sulphate in³⁴S. Those parts of gypsum which are entirely devoid of native sulphur, exhibit a³⁴S value typical of Zechstein (upper Permian). The average enrichment of³²S in native sulphur of about 26% suggests its bacteriological origin, since under the temperatures usually prevailing during the origin of a salt dome, a sulphate cannot be reduced by petroleum. During the rise of the salt dome, the sulphates became contaminated with a small amount of petroleum as a result of which a congenial ecological condition for the development of the sulphate reducing bacterias was produced. The³⁴S depleted H₂S, produced by these bacterial, reacted with the sulphate, thus giving rise to native sulphur (Feely andKulp, 1957).The activity of the sulphate-reducing bacterial leads to different stages of fractionation in the different biochemical environments owing to the varying intensity of the bacterial growth, as well as to the variable ratios of reduced sulphate to primary sulphate. The highly mobile H₂S was homogenized already before the precipitation of native sulphur, whereas the homogenisation of the sulphate is viewed in the solution and recrystallisation processes taking place just below the Neocomian transgression horizon.

     

Finite-element modelling of folding and faulting

The paper is concerned with the finite-element modelling of geologic phenomena. More specifically, the paper presents numerical models of folding and sait doming. A (penalty-) finite-element model based on the equations governing a viscous incompressible fluid is used to simulate the folding is used to simulate the folding of a viscous layer in a less viscous matrix and the salt dome formation, and to predict the location of active faulting and deformation in prograding delta systems on continentalmargins. The initiation and development of drape folding and associated faulting are also modelled. The United States continental margin of the Gulf of Mexico is used as an example of prograding delta system. The finite-element predictions seem to agreewith the published results.