V11A-0358
Insights Into the Mechanisms of Shallow Dike Intrusions Through Particle Code Modelling
Volcanic eruptions are assumed to result from an intruding dike reaching the surface. A conduit is simply the same dike pathway used repeatedly to transport magma over a longer time period. Consequently, understanding the mechanism of dike intrusion at shallow depths is critical to our ability to predict volcanic eruptions. Until recently it was commonly assumed that most dikes were feeders to eruptions, however, recent field based studies have suggested that many more dikes are intruded than ever reach the surface. Almost all previous dike work has focused on the problem of buoyant dike propagation through a combination of analytical studies and gelatine experiments. However, these studies are not generally applicable at shallow depths where buoyancy may play a lesser role. The few shallow depth studies have considered only the static elastic stress state of the country rock prior to intrusion and its impact on the direction of dike propagation. The critical brittle behaviour, the impact of the intrusion itself on the local stress field and the potential damage to the country rock are ignored. This study uses a particle code to investigate intrusion behaviour at shallow depths through weak sandstones and strong granites. Starting from small models calibrated with laboratory tests, an upscaled set of models with dimensions of 500x100m are used to investigate intrusion behaviour under lithostatic, compressional and extensional regimes. The model results show observed field relations such as microfracturing and dike jointing and a scale dependence to fracture toughness. In addition, the intrusions in most of the models do not readily propagate towards the surface. This study is of interest to those in volcanology and also mineral exploration as the evolving microfractures develop permeable pathways for volatile escape from the intrusion.
V11A-0359
Cup-shaped Intrusions, Morphology and Emplacement Mechanism Investigate Through Analogue Modelling
We investigate the morphology of large-scale shallow-depth magma intrusions and sub-volcanic complexes with analogue models. Intrusions of analogue magma are done in a granular material that can contain a ductile layer. The model surface is flat to model the formation of plutonic intrusions and it is overlain by a cone when modelling late sub-volcanic complexes. For flat-top models, we obtain cup-shaped intrusions fed by dykes. Cup-shaped intrusions are inverted-cone like bodies. They are different from saucer-shaped intrusions as they possess neither a well developed sill-base, nor an outer rim. However, like saucers, cups are shallow depth intrusions that dome the country rocks. They initiate from an advancing dyke and first develop an inverted-cone like morphology. Then, the central thickness increases and thrusts form at the edge of the domed country rocks. At this stage, the intrusions progressively involve toward a lopolith shape. By using analogue magma of various viscosities we have been able to constrain key relationships: higher intrusion viscosity causes deeper initiation and the deeper they initiate, the larger is the intrusion diameter. A natural example of such intrusion might by the circles of volcanoes like the Azufre-Lastaria (Peru) that might be overlain be a large-scale cup-shaped intrusion. When adding a cone at the surface of the model and, sometimes, a thin ductile layer in the substratum, the morphology of cup-shaped intrusions vary. Note that the ductile layer of our models is not thick enough to induce the gravitational spreading of the cone. Generally, cup-shaped intrusions are asymmetric in cross section and elliptical in plan view. Their formation creates extension structures in the cone (croissant-shaped rift, straight rift or normal fault) and thrusts in some sectors below the cone. Both types of structures are bordered by strike-slip faults. Cups and saucers share many similarities, but differ probably in the fact that saucers are partially sills that are guided by stratigraphic horizons. However, the basic formation mechanisms may be the same and saucers could be regarded as a special form of cup.
V11A-0360
Magnetic fabric of saucer-shaped sills in the Karoo Large Igneous Province
Magmatic sill intrusions commonly exhibit a saucer geometry in undeformed sedimentary basins and volcanic rifted margins. Current emplacement models are based on the analysis of the intrusion geometry and their spatial relationships with potential feeders, not on the knowledge of the magma flow geometry. The Karoo Basin of South Africa hosts hundreds of saucer-shaped sills. Amongst these, the Golden Valley Sill is well-exposed and displays the connections with adjacent and nested saucers. A combination of detailed fieldwork observations and the anisotropy of magnetic susceptibility measurements were used to identify strain markers that can be interpreted in terms of magma flow directions. A total of 113 localities (6 specimens/site), mostly including opposite sill margins, have been sampled for anisotropy of magnetic susceptibility (AMS) analyses. The magnetic properties were defined by measuring hysteresis cycles and K-T curves on 34 and 19 specimens, respectively. The majority of the localities display well-defined magnetic foliations that consistently dip outward from the centre of the Golden Valley Sill. This orientation of the magnetic foliation most likely represents inflation/deflation cycles of the intruding sill that interacts with non-static enclosing walls. In addition, four magma channels were identified and display an imbrication of the magnetic foliation that indicates an outward magma flow direction. In conclusion, the observed magma flow geometries derived from macroscopic flow indicators and the AMS data correlate well and are used to constrain an emplacement model for the Golden Valley Sill Complex. Finally, the emplacement model of sill complexes repeats the cycle -injection of magma - formation of a saucer-shaped sill - pressure build up - fracturation and pressure drop - channeling of magma - injection of (new batch of) magma - formation of a new saucer-shaped sill- until the magma supply stops. http://folk.uio.no/polteau
V11A-0361
The geometry and emplacement of conical sandstone intrusions
Conical sandstone intrusions with a geometry comparable with that of many igneous sills have been identified using 3D seismic data from large areas of the North Sea and Faeroe-Shetland Basins. These intrusions are of reservoir scale, ranging from 100-2000m or so in diameter, 50-300m in height, and 1-80m in thickness (aperture). They are concentrated in specific stratigraphic intervals in the Cenozoic fills of both basins. Two geometrical end members are recognised and defined here: ‘apical cones' and ‘flat-based bowls.' The former consist of inward dipping conical inclined sheets meeting at a prominent apex and the latter of similarly dipping discordant margins climbing from the margins of a concordant sheet. Both end members are associated with domal folds that are interpreted as resulting from the hydraulic elevation of the overburden during intrusion, and which are analogous to similar structures associated with bowl-shaped igneous sills. Measurements of aperture (ù) versus distance exhibit systematic relationships with the structural relief of these folds, offering a potentially predictive method for estimation of sandstone intrusion aperture and reservoir volume prior to drilling. A growth model for these end-member geometries is presented, drawing on existing theory for igneous sill emplacement. Aperture versus distance plots (ù – X) are used to illustrate two contrasting models for aperture inflation during propagation, but these require much further data before any specific growth model can be adopted.
V11A-0362
Formation of D- and I-shaped geochemical profiles in saucer-shaped sills due to post- emplacement magma flow induced by thermal stresses
There are still unresolved problems in the processes of emplacement and crystallization of saucer shaped sill intrusions. We use geochemistry and numerical modelling in order to constrain identify processes in mafic sill intrusions. Profiles sampled through through a saucer-shaped sill complex in the Karoo igneous province, South Africa show a variety of geochemical variations. Some variations are observed repeatedly, i.e. the D- and I-shaped profiles. D-shaped profiles are recognized by having the least evolved composition in the center (high Mg#) with more evolved composition at the upper and lower margins (low Mg#), resulting in a D-shaped Mg# profile. I- shaped profiles are recognized by having no variation in the Mg# through the profile. The formation mechanism of D-shaped profiles is enigmatic, as classical fractional crystallization theory predicts C-shapes to occur. The least evolved composition will be at the margins where crystallization initiates, and with continued cooling and crystallization the center will be progressively more evolved. Hence, we need another formation mechanism. The most common explanation for D-shaped profiles is a movement of early formed phenocrysts towards the center due to flow segregation. However, petrographical evidences from a D-shaped profile in this study show no phenocryst assemblage in the center, and the modal composition is homogeneous through the profile. We propose that differentiation is caused by a melt flow from the central parts of the sill towards the margins driven by underpressure anomalies at the margins. The underpressures develop because of strong cooling gradients at the margins, assuming no volume change due to a rigid crystal network. The less compatible elements associated with the melt phase will be transported into the margins by advection, resulting in a more evolved total system composition from a higher total melt percentage. The central parts will progressively be depleted in the less compatible elements due to the outflux through a stationary crystal network, resulting in a less evolved total system composition from a total lower melt percentage. Dimensional analysis combined with a numerical model developed using finite element method is used to constrain the post-emplacement melt flow induced by cooling. The numerical model show that large underpressure-anomalies (in the order of -1e8 Pa) develop at the cooling margins. The melt flow following Darcy's law of porous flow is integrated over time into total melt displacement. The two main parameters controlling the magnitude of the melt displacement are permeability of the crystal network and viscosity of the melt. By using a basaltic viscosity and the permeability of 90% crystallinity we get a total melt displacement that is larger than 10% of the sill thickness. We therefore conclude that the post-emplacement flow as a differentiation mechanism causing D-shaped profiles is feasible under natural occurring conditions. I-shaped profiles is predicted to occur where there is no or limited flow due to e.g. rapid cooling (i.e. rapidly decreasing permeability and increasing viscosity) or in melts with high silica contents and thus higher viscosity. This model can be applied to any magmatic sheet-intrusion, regardless of orientation.
V11A-0363
Saucer-shaped sills: occurrences, emplacement and implications
Magmatic sill intrusions commonly exhibit saucer geometries in undeformed sedimentary basins and volcanic rifted margins. Despite their broad occurrences, these saucer-shaped structures are poorly known. The aim of this contribution is to present an overview of the geological settings of saucer-shaped intrusions, present an updated emplacement model and to encourage their identification on Earth and on other planets. Saucer-shaped sills are intimately related to large igneous provinces associated with undeformed sedimentary basins (e.g. South Africa, Siberia, North Sea Basin…). These intrusions can act as water reservoirs (e.g. in the arid South African Karoo Basin) and affect oil maturation and migration pathways. The current emplacement models are based on the analysis of the intrusion geometry and their spatial relationships with potential feeders, not on the knowledge of magma flow directions. New anisotropy of magnetic susceptibility data suggests that saucer- shaped sills are emplaced radially and that the overburden influence the magnetic fabric. Additional analogue and numerical modeling show that saucer-shaped sills are generated by the interactions of a shallow fracture growing radially towards the free-surface and the overburden deformations. A correlation factor of 4-5 between the depth of emplacement and inner diameter exists for saucer-shaped sills. Saucer-shaped intrusions can develop in materials with totally different physical properties (i.e. elastic and Mohr-Coulomb materials) and therefore represent a fundamental geometry in natural systems. We also suggest that saucer-shaped sills should be recognized on extra-terrestrial planets with known intense volcanic activity (e.g. Mars and Venus).folk http://folk.uio.no/polteau
V11A-0364
Dome Structures Above Sills and Saucer-Shaped Sills: Insights From Experimental Modeling
Saucer-shaped magma and sand intrusions are common features in sedimentary basins. They result from fundamental processes for the emplacement of fluids in shallow sedimentary basins. Seismic data show that the overburden above saucer-shaped intrusions is usually deformed and exhibits a dome-like structure. The formation of such structures, and the associated deformation, are of primary importance in the evolution of petroleum systems. In this presentation, we report on experimental investigation of the deformation processes associated with the intrusion of saucer-shaped intrusions into sedimentary basins. The experimental setup consists of molten low-viscosity oil injected into fine-grained silica flour (see Galland et al., this session). It properly simulates the emplacement of saucer-shaped intrusions and the deformation of the country rock. During experiments, the surface of the model is digitalized through a structured light technique based on moiré projection principle. Such a tool provides topographic maps of the model and allows a periodic (every 1.5 s) monitoring of the model surface. When the model magma starts intruding, a symetrical dome rises above the inlet. As injection proceeds, the dome inflates and widens. Subsequently, the dome evolves to a plateau-like feature, with nearly flat surface and steep edges. The plateau keeps lifting up, but nearly stoppes widening. At the end of the experiments, the intruding liquid erupts at the edge of the plateau. The intrusion formed in the experiment is a typical saucer-shaped sill. The evolution of the deforming surface reflects the evolution of the intrusion. We infer that the first doming phase corresponds to the emplacement of a horizontal basal sill by open fracturing. The dome-to-plateau transition corresponds to a transition of the liquid emplacement mechanism from basal sill to inclined sheet. We suggest that the emplacement of the inclined sheets results from shear fracturing at the dome edge.
V11A-0365
Deformation of Magma-Filled Bodies during Solidification
As magma or lava solidifies, volatiles are concentrated in the residual liquid. The result will be expansion (including venting) or pressurization. The mechanism behind this is well-described. A rough hand calculation indicates that an alkali basalt with 4 wt% volatiles would attain attain 12 MPa with 50% crystallization at constant volume. Such pressures would easily be enough to break through the roof of a typical lava tube. If confined in a tunnel deeper in the ground, even in a relatively weak rock, crystallization would be virtually isochoric. However, in a sill at depths of only a few hundred meters, expansion could result in more nearly isobaric crystallization. In either event, before cooling enough to become a brittle solid, the outer portions of the magma would reach a viscoplastic state that could seal in any remaining vapor phase. This would allow pressures to increase further as solidification progressed. Using PELE, a computer code developed to calculate the progress of solidification (Boudreau, 2005), we calculate isochoric and isobaric equilibrium crystallization of alkali basalt and obtain pressures and viscosities as a function of temperature. For an initial pressure of 6 MPa and 0.85 weight percent water, the liquidus is 1433 K. The isochoric pressure reaches 11 MPa at 1293 K with 57% of the mass crystallized; the bulk viscosity is about 3 MPa-s, but that of the residual liquid is only 1 kPa-s. At the same temperature, the isobaric path results in 60% crystallization and a viscosity on the order of 10 kPa-s. A tabular body with these properties would be easily deformed by sagging of the roof if the viscoplastic seal were breached, resulting in a saucer shape. With 91% of the mass crystallized, the isochoric pressure exceeds 28 MPa at 1173 K. By that time, the bulk viscosity of the nearly crystallized mass is on the order of 1025 Pa-s, effectively solid, and the viscosity of the residual liquid (there is also a vapor phase) is about 50 kPa-s. Combining these results with those of a calculation of the cooling of magma in a horizontal cylinder, we find that the interior of the magma is at high-pressure (>10 MPa) and quite fluid (μ < 1 kPa-s) for long enough times to allow considerable deformation of any structures that might be entombed in magma in a deep tunnel. Reference: Boudreau, 2005, http://www.nicholas.duke.edu/people/faculty/boudreau/DownLoads.html
V11A-0366
How Deep Can be a Dyke? II: Common Misconceptions.
The choice of a conceptual model of a dyke either as a continuous conduit joining the region of magma storage and the surface during an eruptive event or, alternatively, as isolated batches of ascending magma, depends on a number of hypothesis that might not be explicitly stated. For instance, it is often assumed 1) that the internal pressure of a dyke equals the pressure exerted by the outer rock at the middle of the dyke, 2) that laboratory measured tensile strengths or fracture toughness of rocks control fracture initiation regardless of confining pressure, or 3) that the mantle behaves as an elastic solid during dyke events. These, and other hypothesis, are justified if fracture conditions are those of linear elastic fracture mechanics (LEFM). Geological and physical evidence, however, suggest that a) mantle rocks are a viscoelastic material with a relaxation time of ca. 34 yrs, b) confining pressure can invalidate the conditions behind LEFM, and c) the strong rheological contrast between magma and the surrounding rock influences the form in which stresses are transmitted in both materials. When these observations are accounted for, not only it becomes evident that LEFM might not be appropriate to describe the formation of dykes, but also it is shown that the continuous conduit model of a dyke is favored. Consequently, the existence of dykes extending to depths of 250 km beneath the surface is a physical reality. These observations also are important in controlling the ultimate 3D aspect of an intrusive, and might play an important role in the development of saucer shaped intrusions.
V11A-0367
FEM Modeling for Magma Chamber of Unzen Volcano Causing 1990-1995 Eruption Inferred From Seismic and Ground Deformation Data
The eruption during 1990~1995 of Unzen Volcano located in Kyushu, Japan, was caused by ascended magma from deep part beneath Chijiwa Bay, lying in the 15 km west from crater of the Volcano. Hypocenters of the earthquake swarm occurring around Unzen area during 1989~1990 were migrated to eastward from deep part of Chijiwa Bay. We propose a model of magma chambers for the eruption using seismic and geodetic observation data. Three types of ground deformation data, leveling, GPS and tidal data were applied in our previous study, and four major point sources beneath Unzen area were founded. The magma sources were arranged as ascending start from 15km depth part beneath Chijiwa Bay toward the crater. Stress distribution around magma source could give us constraint about the shape of magma chamber. Therefore we also performed stress tensor inversion using polarities of first onset of P wave for many earthquakes around the Volcano. The minimum principal stress axes in deeper part (> 15 km) obtained by this method have North-South direction, that is similar with the tension axis of tectonic stress field around Unzen Volcano. On the other hand the axes of maximum principal stress in shallower part of the region have directions toward the conduit of magma, suggesting expansion of conduit occurred from magma intrusion. Based on these geophysical implications we carried out a simulation by Finite Element Method (FEM) to get accurate model during eruption (1990~1995) until present. In previous study we applied semi-infinite media and regarded magma source as point spherical source (Mogi source), in this study we considered topography and shape of magma source obtained by stress distribution. As the result we obtained best-fitted model to the leveling data, the GPS data, and the tidal data around Shimabara peninsula as well as stress data around Unzen Volcano.
V11A-0368
Quantitative Analysis of Circular Symmetry of Venus Coronae and Craters
The origin of craters has long been debated: Exogenic or endogenic? Impact or volcanic? While for the craters of the Earth and Moon the issue has been largely resolved, it has flared anew in recent papers by Hamilton (2005, 2007), Vita-Finzi et al. (2005), and Jurdy and Stoddard (2005, 2007). We weigh in with a quantitative technique to differentiate between these possible mechanisms. Craters by their nature are circular. They are excavated by a roughly hemispherical shock wave, and thus almost regardless of impact angle, will be round rim-and-basin structures (Melosh, 1989). Although underlying structural features, such as faults, and later tectonic deformation can affect crater shape we suggest that the strongest test of an impact origin for coronae is the circularity of these features. Here we introduce an approach for the assessment of a feature's circular symmetry. Using altimetry data we compare, by cross-correlation, multiple profiles across a single feature. Jurdy and Stoddard (2005) provided an example in which Mead crater and two coronae were analyzed. They found that for each corona, profiles cross- correlated at only 25-30% of perfect cross-correlation. Profiles for Mead crater, however, correlated at a much higher level, 80%. Here, we perform an expanded study of features generally classified as craters, and others whose classification as coronae has been questioned by Hamilton (2007). We choose only the largest craters, since altimetry data are too coarse to allow enough data points for analyses of smaller features, and also because they are of similar size to the coronae in our study. For each feature, 36 profiles are extracted from the altimetry data, de-sloped, and averaged together. For each feature, the individual profiles are correlated against the average, and the correlations themselves were averaged to give an assessment of circular symmetry. Results indicate accepted craters have the highest correlation averages (are most circular) and suggest that a few coronae may actually be of impact origin. We propose that this type of correlation analysis can be used in an objective assessment of circularity, and therefore the origin, of the remaining catalogue of similar features.
V11A-0369
Scaling Relationships for Deep and Shallow Magma-Driven Intrusions
Components of the elastic fracture/lubrication model have been used in the past to evaluate the growth of magmatic intrusions, but it has remained unclear whether this model is capable of producing results which are consistent with field data. The mathematical model used in this study accounts for full coupling between fluid flow and the elastic deformation and fracture of the rock as well as the mechanical influence of the Earth's surface on shallow intrusion growth. A scaling method is used to derive approximate relationships between thickness and length for circular intrusions growing in each of four propagation regimes that correspond to deep and shallow cases which can be either viscosity or toughness dominated. Application of these results to particular field cases requires careful consideration of specific governing parameters, geometry, and other local conditions. In contrast, this work relies on choosing governing parameters such as magma viscosity and injection rate from published ranges and then employing the derived relations to bound the model's solution domain using a Monte-Carlo technique. In this way the viability of the model is examined in broad terms in order to understand its applicability to modeling magmatic intrusions in general. The results indicate that the solution domain corresponds well with available data both in the values of the estimated thickness to length ratios and in the tendency for small/deep intrusions to favor lateral spreading while large/shallow intrusions favor vertical inflation or growth with a fixed thickness to length ratio.
V11A-0370
Saucer-shaped sills at shallow depths beneath a scoria cone volcano (Paiute Ridge, Nevada)
Late Miocene sills and dikes in the Paiute Ridge area of southern Nevada were emplaced in an extensional setting beneath a small volume, alkali basaltic volcano. Dikes (400-5000 m long, 1.2-9 m wide) mostly occupy pre-existing E-dipping normal faults. Three small sills (extending laterally up to ~500 m, and 20-46 m thick) and two larger sills (each having lateral dimensions ~1 km) locally branch off some dikes within ~250 m of the paleosurface. Individual small sills extend only into the hanging wall blocks of the faults that host their parent dikes, and are connected to the dikes by stems that are only a few tens of meters wide; elsewhere along their strikes the parent dikes extend above the sills. We infer that sill emplacement was caused by local rotation of principal stresses related to the intersection of the dike-hosting fault planes with the complex contact between relatively strong Paleozoic carbonates and weak Tertiary tuffs. This effect might have been accentuated by co- intrusive displacement along the faults as they were lubricated by magma, an interpretation that is corroborated by hydromechanical calculations. Orientation of bedding planes in the tuffs controlled the direction of sill propagation. The three most areally extensive sills formed lopoliths with sagging roofs, indicating interaction with the free surface. Both the bedding plane and free surface interactions contributed to the dish-shaped geometry of some of the sills (Ref: Earth Planet. Sci. Lett. 246, 217-230, 2006).
V11A-0371
The Significance of Fingers and Lobes in Emplacement of Sill Complexes: Insights From Field Data
The understanding of the emplacement of sill complexes has developed greatly of the last decade from insights gained from 3D seismic data. Particularly, the identification of structures, such as units and lobes in the order of 100s m - km scale, which make up individual sills (e.g. Thomson & Hutton, 2004, Bull Volcanol, 66, 364–375). However the formation and significance of these structures in terms of the emplacement of sill complexes is only partially understood. We present field data from planar-segmented sills on the Isle of Skye, Scotland and saucer- shaped sills in the Karoo Basin, S Africa. In both of these areas constituent lobes can be confirmed but we have also observed smaller scale finger-like structures in the order of 10s – 100s m, which appear to make up these lobes. In planar-segmented sills (Skye) fingers are aligned along the axis of the lobes and have a NNW trend, consistent with a structurally controlled emplacement direction. In saucer-shaped sills (Karoo) with no obvious structural control, fingers are oriented radially about lobes, radiating outwards from the centre of the saucer. Fingers initially propagate separately, thickening vertically and laterally and coalescing to form lobes. Once the fingers have coalesced, the sill continues to thicken vertically. Finger structures clearly have a fundamental role in the initial propagation of sills. Detailed examination of fingers of the Golden Valley Sill in the Karoo Basin shows that there is a regular and predictable spacing between the crests and troughs between individual fingers. This spacing is consistent within individual lobes but varies from lobe to lobe. This implies that sills are constructed in a stepwise fashion and that models which assume one discrete intrusion event may be an oversimplification. The regular spacing of fingers also leads us to suspect that their wavelength may provide information about the propagation of the sill. It is proposed that the wavelength of the fingers of an individual lobe may be a function of the velocity and viscosity of the magma. According to our theory, shorter wavelength (narrow) fingers are produced by relatively low viscosity and higher velocity magmas, indicating that the sill front propagated relatively quickly. Longer wavelength (thicker) fingers result from higher viscosity, lower velocity magma, indicating that the sill propagated relatively slowly.
V11A-0372
Geochemical Architecture of the Golden Valley Sill Complex, South Africa: Implication for the Emplacement of Saucer-Shaped Sills in Sedimentary Basins
Saucer-shaped sills and dykes are common features in sedimentary basins worldwide and represent important parts of the plumbing system in such areas. In spite their common occurrence, the emplacement mechanisms that lead to the formation of sill complexes are poorly understood. Two main emplacement models have been proposed, based on field observations, seismic imaging, and mathematical modeling: (1) Each sill in a complex is produced by an individual magma batch emplaced through a dyke or a pipe, ending in the formation of a sill. (2) A single batch of magma gives rise to a series of saucer-shaped sills which feed one another, thereby forming a nested sill complex in which the different sills are interconnected. We report on a detailed geochemical study of a complex of nested, saucer-shaped dolerite sills, the Golden Valley Sill Complex in the Karoo Large Igneous Province, South Africa. This well-exposed sill complex consists of four large sills (ca. 100 m thick; long axes: 13- 24 km) emplaced at slightly different stratigraphic levels, one small sill (55-80 m thick; long axis: 4 km; forming an appendix to one of the large sills), and two large dykes (15-20 m thick; 25 and 70 km long). The field observations show no physical connections between the large sills, or between the sills and the dykes. The compositions of basalts and dolerites in the Karoo basin are very similar. However, small chemical differences do exist. In order to discriminate between sills and dykes of different geochemical signature in the Golden Valley Sill Complex, that might have formed from different magma batches, we used a Forward Stepwise-Discriminant Function Analysis (FS-DFA). Sample groups which, based on field observations, were found clearly to belong to the same sill or dyke were defined as populations, the variables were forty-seven major and trace elements. The FS-DFA showed that several magma batches of distinct chemical characteristics were involved in the formation of the GVSC. Four different magma batches gave rise to two of the large sills and the two dykes. The exposed dykes thus do not represent feeders to the main sills in the GVSC. One additional, chemically distinct magma batch gave rise to the two other large sills plus the small sill. Our study of the GVSC thus gives support to both the main models proposed earlier to explain the emplacement of sill complexes. Some sills in a sill complex may be fed by different magma batches (model 1); other sills in the same complex may overflow and feed new sills, forming a group of nested sills formed from one single batch of magma (model 2). An important difference between model (2) and our observations is the way sills feed one another. Model (2) proposes that each of the nested sills in a complex is fed from its center, connecting the distal part of one sill to the center of the next sill which floor will be located at a higher stratigraphic level. In the GVSC the nested sills are connected along their edges through lateral overflow, the nested sills are thus located at approximately the same stratigraphic level. The GVSC includes complex areas where two or more sills appear to meet, and where two sills are located above one another. Using the FS-DFA parameters for the main populations, we were also able to assign samples in these areas to specific sills, thus improving the geological mapping of the area.
V11A-0373 INVITED
Saucer-shaped Clastic Intrusions and Associated Injectites in the Westerm San Joaquin Valley
Clastic sills, including saucer-shaped intrusions, are the most volumetrically significant clastic intrusions in the Panoche Giant Injectite Complex (PGIC). Injection occurred in the Lower Paleocene during a period of inversion caused by the convergence of the Pacific and North American plates. Almost 400 km2 of exposure reveals the relationships between clastic intrusions, their parent beds and seafloor sand extrusions (extrudites). Sand was injected into partially-consolidated deepwater mudstones of late Cretaceous and early Paleocene age in a single event but with many pulses that produced cross-cutting intrusions. More than 40 km3 of sand is estimated to have injected within the area of outcrop. The total thickness of strata cut by injections (from deepest known Lower Cretaceous parent bed to extrudite) is in excess of 1.2 km. Saucer-shaped intrusions are composite features that comprise sills, low-angle dikes and dike swarms, arranged as low-angle conical or saucer-shaped injected units that exhibit a semi-elliptical to horse-shoe geometry in plan view and are V- or U-shaped in cross section. Saucer-shaped intrusions are 500 to 1.5km wide and in some cases cut through more than 250 m of stratigraphic section. The lowest parts of the saucer-shaped intrusions consists of multiply-stacked, low-angle dikes up to 80m thick that cut the stratigraphy at angles between 5-10º and locally exhibit stepped or transgressive geometry. The low-angle dykes are 8 to 20 m wide and laterally continuous over distances of a few to several hundreds of metres. Steeper segments (up to 30 deg) emanate laterally from the periphery of the lowest units and cut through the host-rock for a few up to several tens of metres, pinching-out laterally over distances of several tens to a few hundreds of metres. At their margins the saucer-shaped intrusions are bounded by steep (50-70 deg), narrow dikes (generally <1 m wide); such dikes are unlikely to be imaged on seismic data. The central areas of the saucer-shaped intrusions contain uncommon, narrow, widely spaced, thin dikes. Evidence for forced folding and jacking up of the host-rock related to emplacement of the saucer-shaped intrusions is identified Multiple dikes terminate at the base of the saucer- shaped intrusions; the dikes have variable apertures ranging from a few centimetres up to 2 m and are inferred to be feeders of the saucer-shaped intrusions. http://www.abdn.ac.uk/geology/people/staffpages/hurst/hurst.php#si
V11A-0374
Experimental Modeling of the Formation of Saucer-Shaped sills
Many magma intrusions in sedimentary basins are sills, and especially saucer-shaped sills. These features are observed in many places (i.e. South Africa; the Norwegian and North Sea; Siberia; Argentina). Sand injectites exhibit similar geometries. The occurrence of such features in so various settings suggests that their emplacement results from fundamental processes in sedimentary basins. To understand such processes, we performed experimental modeling of saucer-shaped sill emplacement. The experiments consist of injecting a molten low viscosity vegetable oil (model magma) at a constant flow rate into a fine-grained Coulomb silica flour (model rock). When the oil starts intruding, the initially flat surface of the model inflates and forms a smooth dome. At the end of the experiment, the oil erupts at the edge of the dome. After the experiment, the oil cools and solidifies, the resulting solid intrusion is unburied and exposed, and its upper surface digitalized. For our purpose, we did our experiments without external deformation. We performed two series of experiments with varying depth of injection. The first series consisted of injection into a homogeneous medium. The resulting intrusions were cone-sheets and dykes. The second series consisted of heterogeneous models where the heterogeneity was a weak layer made of a flexible net. The resulting intrusions were made of (1) a horizontal basal sill emplaced along the weakness, and (2) inclined sheets nucleating at the edges of the basal sill and propagating upward and outward. The inclined sheets exhibited a convex shape, i.e. a decreasing slope outward. In addition, the deeper the sills emplaced, the larger they were. Our experimental results are consistent with saucer-shaped features in nature. We infer from our results that the transition between the basal sills and the inclined sheets results from a transition of emplacement processes. We suggest that the basal sill emplace by open (mode I) fracturing, whereas the inclined sheets result from shear (mode II) fracturing, i.e. along faults at the edge of the dome.
V11A-0375
Saucer-Shaped Sandstone Intrusions: Facts, Inferences and Unknowns
Saucer-shaped and conical sandstone intrusions occur in abundance within Paleogene claystones of the North Sea Basin and within Paleogene-Neogene claystones of the Faroe-Shetland and More Basins along the NW European Atlantic Margin. The dimensions of individual saucers range from 50-300m height, 0.5-2 km width, and 0.5-4 km length, with sandstone volumes up to some 0.5 cubic kilometres. Clusters of saucers may contain composite intruded volumes up to several cubic kilometres and may form significant reservoir bodies for hydrocarbon accumulations. Conical sandstone intrusions have similar dimensions, though their width, length and volumes are limited by their downward tapering geometry with a central pointy apex, lacking a horizontal segment in the centre. Whilst their occurrence, dimensions and significance within their known host basins are no longer in doubt, virtually all the parameters relating to their genesis are poorly constrained and either inferred or declared unknown in the existing literature. These include: source of the intruded sand, timing of intrusion (both duration and age), depth of emplacement, triggering mechanism(s), relation with underlying structures and/or structures within and the rheology of the host claystones, etc. It is also largely unknown whether intrusions occurred largely synchronously or during a multitude of events within their host basins. Without a rigorous analysis of which parameters are known and which are inferred or essentially unknown, any analysis of the origin of sandstone intrusions remains speculation and may be highly misleading. This paper presents examples and highlights the facts, inferences and unknowns for each of the case studies drawn from the northern North Sea and the Faroe-Shetland Basin and summarises the implications of these uncertainties for the analysis of the origin of the intrusions.