Clustering and visualization of earthquake data in a grid environment

We present a web client-server service WEB-IS, which we have developed for remote analysis and visualization of seismic data consisting of both small magnitude events and large earthquakes. We show that the problem-solving environment (PSE) intended for prediction of large magnitude earthquakes can be based on this WEB-IS idea. The clustering schemes, feature generation, feature extraction techniques and rendering algorithms form a computational framework of this environment. On the other hand, easy and fast access both to the seismic data distributed among distant computing resources and to computational and visualization resources can be realized in a GRID framework. We discuss the usefulness of NaradaBrokering (iNtegrated Asynchronous Real-time Adaptive Distributed Architecture) as a middleware, allowing for flexibility and high throughput for remote visualization of geophysical data. The WEB-IS functionality was tested both on synthetic and the actual earthquake catalogs. We consider the application of similar methodology for tsunami alerts.     

A model for oxygen and sulfur isotope fractionation in sulfate during bacterial sulfate reduction processes

We present a model of bacterial sulfate reduction that includes equations describing the fractionation relationship between the sulfur and the oxygen isotope composition of residual sulfate (δ³⁴S_{SO4_residual}, δ¹⁸O_{SO4_residual}) and the amount of residual sulfate. The model is based exclusively on oxygen isotope exchange between cell-internal sulfur compounds and ambient water as the dominating mechanism controlling oxygen isotope fractionation processes. We show that our model explains δ³⁴S_{SO4_residual} vs. δ¹⁸O_{SO4_residual} patterns observed from natural environments and from laboratory experiments, whereas other models, favoring kinetic isotope fractionation processes as dominant process, fail to explain many (but not all) observed δ³⁴S_{SO4_residual} vs. δ¹⁸O_{SO4_residual} patterns. Moreover, we show that a “typical” δ³⁴S_{SO4_residual} vs. δ¹⁸O_{SO4_residual} slope does not exist. We postulate that measurements of δ³⁴S_{SO4_residual} and δ¹⁸O_{SO4_residual} can be used as a tool to determine cell-specific sulfate reduction rates, oxygen isotope exchange rates, and equilibrium oxygen isotope exchange factors. Data from culture experiments are used to determine the range of sulfur isotope fractionation factors in which a simplified set of equations can be used. Numerical examples demonstrate the application of the equations. We postulate that, during denitrification, the oxygen isotope effects in residual nitrate are also the result of oxygen isotope exchange with ambient water. Consequently, the equations for the relationship between δ³⁴S_{SO4_residual}, δ¹⁸O_{SO4_residual}, and the amount of residual sulfate could be modified and used to calculate the fractionation-relationship between δ¹⁵N_{NO3_residual}, δ¹⁸O_{NO3_residual}, and the amount of residual nitrate during denitrification.     

Earth's rotation variations: a wavelet analysis

Historical to up‐to‐date data of the minute variations in the solid Earth's rotation are subjected to a comprehensive time‐frequency wavelet analysis. The length‐of‐day for the period 1962–2012 confirms the presence of a prominent, robust 6‐year periodicity and reveals an anomalously strong 18.6‐year tidal oscillation as well as a ~13‐year quasi‐periodic signal. The polar‐motion excitation for the period 1900–2012 validates the existence of the ~26‐year Markowitz wobble and shows an ~8‐year quasi‐periodic signal, but no appreciable 18.6‐year periodicity. Although it is known that exchanges of angular momentum with the geophysical fluids are responsible for the rotational variations of the solid Earth, the exact physical mechanisms involved on interannual‐to‐decadal timescales are still far from understood.     

Testing the suitability of geologic frameworks for extrapolating hydraulic properties across regional scales

The suitability of geologic frameworks for extrapolating hydraulic conductivity (K) to length scales commensurate with hydraulic data is difficult to assess. A novel method is presented for evaluating assumed relations between K and geologic interpretations for regional-scale groundwater modeling. The approach relies on simultaneous interpretation of multiple aquifer tests using alternative geologic frameworks of variable complexity, where each framework is incorporated as prior information that assumes homogeneous K within each model unit. This approach is tested at Pahute Mesa within the Nevada National Security Site (USA), where observed drawdowns from eight aquifer tests in complex, highly faulted volcanic rocks provide the necessary hydraulic constraints. The investigated volume encompasses 40 mi³ (167 km³) where drawdowns traversed major fault structures and were detected more than 2 mi (3.2 km) from pumping wells. Complexity of the five frameworks assessed ranges from an undifferentiated mass of rock with a single unit to 14 distinct geologic units. Results show that only four geologic units can be justified as hydraulically unique for this location. The approach qualitatively evaluates the consistency of hydraulic property estimates within extents of investigation and effects of geologic frameworks on extrapolation. Distributions of transmissivity are similar within the investigated extents irrespective of the geologic framework. In contrast, the extrapolation of hydraulic properties beyond the volume investigated with interfering aquifer tests is strongly affected by the complexity of a given framework. Testing at Pahute Mesa illustrates how this method can be employed to determine the appropriate level of geologic complexity for large-scale groundwater modeling.     

Influence of the Mississippi River on <Emphasis Type="Italic">Pseudo-nitzschia</Emphasis> spp. Abundance and Toxicity in Louisiana Coastal Waters

The presence of domoic acid (DA) toxin from multiple species of Pseudo-nitzschia is a concern in the highly productive food webs of the northern Gulf of Mexico. We documented the Pseudo-nitzschia presence, abundance, blooms, and toxicity over three years along a transect ～100 km west of the Mississippi River Delta on the continental shelf. Pseudo-nitzschia were present throughout the year and occurred in high abundances (>10⁴ cells l^?1) in the early spring months during high Mississippi River (MSR) flow (～20,000 m³ s^?1) but were most abundant (>10⁶ cells l^?1) when MSR discharge was relatively lower among the spring months. A high particulate toxin production (maximum reaching 13 μg DA l^?1) was associated with the high cell abundances and exceeded, by an order of magnitude, prior reports of particulate DA concentrations in Louisiana coastal waters. Differences in Pseudo-nitzschia peak times and its toxicity were correlated mainly with the timing and magnitude of MSR discharge and changes in associated parameters such as nutrient stoichiometry and salinity. A negative relationship between high MSR discharge and Pseudo-nitzschia and particulate DA concentrations was documented. These riverine dynamics have the potential to influence DA contamination in pelagic and benthic food webs in the coastal waters of the northern Gulf of Mexico.     

Precursory slope distress prior to the 2010 Mount Meager landslide,British Columbia

In 2010, the south flank of Mount Meager failed catastrophically, generating the largest (53 ± 3.8 × 10⁶ m³) landslide in Canadian history. We document the slow deformation of the edifice prior to failure using archival historic aerial photographs spanning the period 1948–2006. All photos were processed using Structure from Motion (SfM) photogrammetry. We used the SfM products to produce pre-and post-failure geomorphic maps that document changes in the volcanic edifice and Capricorn Glacier at its base. The photographic dataset shows that the Capricorn Glacier re-advanced from a retracted position in the 1980s then rapidly retreated in the lead-up to the 2010 failure. The dataset also documents 60 years of progressive development of faults, toe bulging, and precursory failures in 1998 and 2009. The 2010 collapse was conditioned by glacial retreat and triggered by hot summer weather that caused ice and snow to melt. Meltwater increased pore water pressures in colluvium and fractured rocks at the base of the slope, causing those materials to mobilize, which in turn triggered several secondary failures structurally controlled by lithology and faults. The landslide retrogressed from the base of the slope to near the peak of Mount Meager involving basement rock and the overlying volcanic sequence. Elsewhere on the flanks of Mount Meager, large fractures have developed in recently deglaciated areas, conditioning these slopes for future collapse. Potential failures in these areas have larger volumes than the 2010 landslide. Anticipated atmospheric warming over the next several decades will cause further loss of snow and glacier ice, likely producing additional slope instability. Satellite- and ground-based monitoring of these slopes can provide advanced warning of future landslides to help reduce risk in populated regions downstream.     

Petrogenesis of low-δ<Superscript>18</Superscript>O quartz porphyry dykes,Koegel Fontein complex,South Africa

This paper investigates the origin of low-δ¹⁸O quartz porphyry dykes associated with the 144–133 Ma Koegel Fontein Igneous Complex, which was intruded during the initial phase of breakup of Africa and South America. The 25-km diameter Rietpoort Granite is the largest and youngest phase of activity, and is roofed by a 10-km diameter pendant of gneiss. Quartz porphyry (QP) dykes, up to 15 m in width, strike NW–SE across the complex. The QP dykes that intruded outside the granite have similar quartz phenocryst δ¹⁸O values (average 8.0‰, ± 0.7, n?=?33) to the granite (average 8.3?±?1.0, n?=?7). The QP dykes that intruded the roof pendant have quartz phenocrysts with more variable δ¹⁸O values (average 1.6‰, ± 2.1, n?=?55). In some cases quartz phenocrysts have δ¹⁸O values as low as ? 2.5‰. The variation in δ¹⁸O value within the quartz crystal population of individual dykes is small relative to the overall range, and core and rim material from individual quartz phenocrysts in three samples are identical within error. There is no evidence that quartz phenocryst δ¹⁸O values have been affected by fluid–rock interaction. Based on a ?_quartz?magma value of 0.6‰, magma δ¹⁸O values must have been as low as ? 3.1‰. Samples collected along the length of the two main QP dykes that traverse the roof pendant have quartz phenocryst δ¹⁸O values that range from +?1.1 to +?4.6‰, and ? 2.3 to +?5.6‰, respectively. These δ¹⁸O values correlate negatively (r = ? 0.96) with initial ⁸⁷Sr/⁸⁶Sr, which can be explained by the event that lowered δ¹⁸O values of the source being older than the dykes. We suggest that the QP dykes were fed by magma produced by partial melting of gneiss, which had been variably altered at high temperature by ¹⁸O-depleted meteoric water during global glaciation at ~?550 Ma. The early melts had variable δ¹⁸O value but as melt pockets interconnected during melting, the δ¹⁸O values approached that of average gneiss. Variable quartz phenocryst δ¹⁸O values in the same dyke can be explained by vertical emplacement, at variable rates of ascent along the dyke. The lateral variation in quartz, and hence magma δ¹⁸O value at a particular point along a single dyke would depend on the rate of ascent of magma at that point along the dyke, and the ‘age’ of the particular magma batch.