A RAM-economizing domain decomposition technique for regional high-resolution groundwater simulation

A major limitation of the use of high-resolution groundwater models on a regional scale for resource management by water companies is the excessive RAM requirements of such models which surpass the capacity of today's PCs. A strategy is presented to overcome this problem by using overlapping domain decomposition techniques. Furthermore, because very long computing time is the bottleneck for the practical use of this technique for large groundwater models, an analysis is also presented of a number of methods implemented to increase calculation performance. The approach presented here is characterized by a fairly simple structure that represents a generalized relaxation algorithm. It can be adapted for use with finite element as well as with finite difference methods. Electronic Publication     

Découverte de structures périglaciaires (sand-wedges et composite-wedges) sur le site de stockage de déchets radioactifs de l'Aube (France)

Since 1990, a dense network of subvertical cracks has been observed at the upper part of the Albian sands of the site of radioactive waste storage operated by ANDRA. New studies carried out in 1999 during enlargement of the site of Soulaines and in 2003 on the nearby TFA site highlighted an especially well-developed network of periglacial features that are mainly of the sand-wedge type (subvertical stratified infill). These observations demonstrate the periglacial origin of these cracks, which are mainly superimposed on a dense network of former tectonic extension joints, and that rapidly disappear at the top of the underlying clays. In addition, these observations currently represent the only example of a large sand-wedge network developed during the Last Glacial in northern France. To cite this article: P. Antoine et al., C. R. Geoscience 337 (2005).     

The Fe/ΣFe ratios of MORB glasses and their implications for mantle melting

The Fe³⁺/ΣFe ratio of 104 MORB glasses from the Pacific, the Atlantic, the Indian, and the Red Sea spreading centers have been determined using wet chemical Fe²⁺ analyses and electron microprobe FeO_total measurements. The data provide a new estimate for the MORB oxygen fugacity (fO₂) of 0.41 ± 0.43 (1sigma, N = 100) log units below the fayalite-magnetite-quartz buffer (FMQ), equivalent to a Fe³⁺/ΣFe = 0.12 ± 0.02 (1sigma, N = 104). This new fO₂ estimate is 0.8 log units more oxidized than the average fO₂ proposed by Christie et al. (1986) (FMQ-1.20 ± 0.44; Fe³⁺/ΣFe = 0.07 ± 0.01; N = 87). This slight difference may be related in part to the 3.5% underestimation of the Fe²⁺ concentration determined by Christie et al. (1986) compared with this study. MORB oxygen fugacity does not display any significant difference between the three main oceanic domains, or between enriched and depleted MORB. Yet, the iron red-ox state ratio shows a broad increase during fractional crystallization. Detailed study of magmatic suites highlights the lack of systematic Fe³⁺/ΣFe ratio fractionation during differentiation. Despite the large variations of inferred partial melting degrees (from 5 to 20%), the present data set does not provide any evidence of Fe³⁺/ΣFe relationships with partial melting proxies such as Na_8.0.Based on the Fe³⁺ systematics during partial melting, it is suggested that the oxidation state of MORB reflects a “buffered mantle melting process” resulting in the apparent compatible behavior of Fe³⁺ during partial melting, and in the relatively constant Fe³⁺/ΣFe ratio irrespective of the extent of melting. This result implies that partial melting processes may be open relative to oxygen. We propose a model where the Fe³⁺/ΣFe ratio in the melt is buffered during partial melting. The MORB Fe₂O₃ systematics can be accounted for by using a fO₂ of FMQ-1 that is equivalent to the average fO₂ reported for abyssal peridotites.     

The meta-igneous granulite xenoliths from Kerguelen Archipelago: evidence of a continent nucleation in an oceanic setting

Xenoliths of meta-igneous origin occur as one of the two main types of ultramafic and mafic xenoliths entrained by alkaline lavas of the Kerguelen islands. These are designated type II xenoliths and are subdivided into three mineralogical groups. Subtype IIa and IIc xenoliths are interpreted as crystallisation products of basaltic melts that were emplaced near the crust-mantle boundary during the early tholeiitic–transitional magmatic activity of the Kerguelen islands. Younger magmatism became more alkaline and subtype IIb xenoliths were formed as high-pressure alkaline cumulates related to the last alkaline volcanic stage. Subsequently, the plagioclase-bearing type II rocks have been re-equilibrated under granulite facies conditions. This addition of mafic material around the crust-mantle boundary is consistent with seismic evidence for crustal thickening to 14–20 km. Calculated compressional seismic velocities (Vp) for the basic granulites are consistent with the range of observed Vp in the low-velocity region underlying the oceanic crust. Such growth in the thickness of the oceanic crust may be caused by intrusion of basalts at different levels in the lithosphere and may provide the heat responsible for granulitic metamorphism in the oceanic setting. This study suggests that basic granulites can account for the observed seismic characteristics of oceanic plateaux and can be important components of Kerguelen oceanic lithosphere where there has been large-scale magma production. Moreover we speculate that the Kerguelen islands and perhaps the surrounding plateau represent a continental nucleation process. Received: 30 September 1997 / Accepted: 17 June 1998     

Cleavage mechanism and surface chemical characterization of phengitic Muscovite and Muscovite as constrained by X-Ray Photoelectron Spectroscopy

Freshly cleaved (001) natural muscovite was chemically characterized by X-ray Photoelectron Spectroscopy (XPS). The investigated muscovite crystals have composition ranging from pure end member to phengitic muscovite. XPS provides experimental evidence that the chemical composition of the near-surface region differs from the bulk. This difference is not due to analytical problems, but to extreme surface sensitivity of XPS. Depth profiling was also carried out after tilting of 60° or 70° of the sample along an axis perpendicular to [001]. After tilting of phengitic muscovite, the concentration of Al increases and of K decreases, and therefore Al is the topmost monolayer. Repulsion between the octahedral and tetrahedral sheets, stacking faults or more probably interlayered phases (gibbsite-like) are responsible for the cleavage. Possible perturbations of surface chemistry due to X-ray radiation and Ar⁺ ion sputtering have also been investigated. X-ray radiation is negligeable, but Ar⁺ ion sputtering introduces important structural modifications, perturbation of the chemical environment of elements and minor chemical variations. Received: 13 November 1996 / Revised, accepted: 21 March 1997     

Seismic hazard assessment for low-seismicity areas — Case study: Northern Germany

The problem of assessing seismic hazard in low-seismicity areas becomes obvious in many practical applications. A typical low-seismicity area, which experienced damaging earthquakes in historical times, is the North German Plain, for which a case study is presented. It is shown how seismic hazard assessments are influenced by different interpretations of historical key earthquakes, changes in b-value as well as variations of the upper bound magnitude assumed for the seismic source regions. The latter strongly influences the hazard results in the case of very low b-values for long return periods.     

Pollen diagram from the Nebraska Sandhills and the age of the dunes

Radiocarbon dates of organic alluvium beneath as much as 40 m of dune sand along the Dismal River have led to the suggestion that the Nebraska Sandhills date from the Holocene rather than the last glacial period. On the other hand, the basal layers of lake and marsh deposits in interdune depressions at three localities date in the range of 9000 to 12,000 yr B.P., implying a pre-Holocene age for the sand dunes. A pollen diagram for one of these sites, Swan Lake, indicates prairie vegetation throughout the last 9000 yr, with no suggestion that the landscape was barren enough to permit the shaping of the massive dunes characterizing the area. Sand was not transported across the site during the Holocene, either during the marsh phase, which lasted until 3700 yr B.P., or during the subsequent lake phase. The sand that buries the alluvium along the Dismal River may represent only local eolian activity, or it may indicate that the younger of the two main dune series identified by H. T. U. Smith (1965, Journal of Geology73, 557–578) is Holocene in age, and the older one Late Wisconsin in age.     

Chi-square regression for seismic strength parameter relations, and their uncertainties, with applications to an M w based earthquake catalogue for central, northern and northwestern Europe

A generalized chi-square regression approach to establishempirical relations between different types of seismic strengthparameters with uncertainties in all input data is presented anddiscussed in comparison with standard least-squares techniques.The chi-square technique can consider errors of individual entriesbut can also be applied when errors are not exactly known and onlyweaker quantitative constraints can be made. It can preserve thesymmetry of the derived relations and is preferred for complexregression models. Results for three types of regression modelsare presented for (1) a linear relation between M _Sand m _bfor events in the North Atlantic Ocean; (2) a quadratic relationbetween M _w and M _L forevents in central Europe; (3) linearrelations between M _L and I ₀,with logarithmic dependency ofthe focal depth, for several regions in central and northernEurope.