Large-scale stochastic linear inversion using hierarchical matrices

Stochastic inverse modeling deals with the estimation of functions from sparse data, which is a problem with a nonunique solution, with the objective to evaluate best estimates, measures of uncertainty, and sets of solutions that are consistent with the data. As finer resolutions become desirable, the computational requirements increase dramatically when using conventional solvers. A method is developed in this paper to solve large-scale stochastic linear inverse problems, based on the hierarchical matrix (or ? ² matrix) approach. The proposed approach can also exploit the sparsity of the underlying measurement operator, which relates observations to unknowns. Conventional direct algorithms for solving large-scale linear inverse problems, using stochastic linear inversion techniques, typically scale as ??(n ² m+nm ²), where n is the number of measurements and m is the number of unknowns. We typically have n ? m. In contrast, the algorithm presented here scales as ??(n ² m), i.e., it scales linearly with the larger problem dimension m. The algorithm also allows quantification of uncertainty in the solution at a computational cost that also grows only linearly in the number of unknowns. The speedup gained is significant since the number of unknowns m is often large. The effectiveness of the algorithm is demonstrated by solving a realistic crosswell tomography problem by formulating it as a stochastic linear inverse problem. In the case of the crosswell tomography problem, the sparsity of the measurement operator allows us to further reduce the cost of our proposed algorithm from ??(n ² m) to $\mathcal {O}(n^{2} \sqrt {m} + nm)$ . The computational speedup gained by using the new algorithm makes it easier, among other things, to optimize the location of sources and receivers, by minimizing the mean square error of the estimation. Without this fast algorithm, this optimization would be computationally impractical using conventional methods.     

Evaluating prey capture by larval mummichogs (Fundulus heteroclitus) as a potential biomarker for contaminants

We evaluated larval prey capture as a "behavioral biomarker" of contamination by examining feeding behavior of larval mummichogs (Fundulus heteroclitus) from many different sites, including a severely contaminated "Superfund" site, moderately contaminated sites, and reference areas. Prey capture ability was related to sediment contaminant levels. The levels of contaminants at a site were highly correlated with each other, so that the impact of individual contaminants was confounded. The number of captures of brine shrimp by mummichog larvae from all sites was highly variable, but significant negative correlations of prey capture were seen with mercury, lead, zinc, cadmium, and PCBs. As observed previously with adults, polyaromatic hydrocarbons (PAHs) did not appear to impair prey capture ability. The only site in which prey capture rates of 8-day old larvae were severely affected was the most highly contaminated Superfund site, Berry's Creek, NJ. This implies that larval prey capture is not as sensitive a behavioral biomarker for contamination as adult behavior studied previously.     

A rapid change in methylmercury tolerance in a population of killifish,Fundulus heteroclitus,from a golf course pond

Embryos of killifish (Fundulus heteroclitus) from Bullhead Bay, Southampton, New York, showed considerable variation in tolerance to methylmercury (meHg) in 1980, 1981 and 1983, in that some females produced eggs that were tolerant to the teratological effects of 0·05 ppm, while other females produced susceptible eggs. However, in 1982, over 40% of the females produced clutches of non-viable eggs (which could not be tested for meHg tolerance) and those fish which produced viable eggs generally produced tolerant ones. After hatching, the larvae were also more tolerant. Accompanying the shift in meHg tolerance was a trend of increasing fin ray count of the females, a parameter which has previously been correlated with the production of more tolerant embryos. In 1982, a nearby population exhibited a distribution of embryonic and larval tolerance and fin ray count comparable with that of the first population in the other years.We hypothesize that the unusually high rainfall in June, 1982 caused an inflow of pesticides and heavy metals from a golf course adjacent to the first site, and that the contaminants were responsible for the striking changes in reproductive success and meHg tolerance. Residues of chlorinated hydrocarbon insecticides were found in fish of the first, but not of the second, population. The variability in the original population may have permitted a segment of it to withstand the inflow of pollutants, while the susceptible individuals produced non-viable eggs. This may have been the cause of the rapid shift of overall tolerance in the population.     

A chemical investigation of the transport and fate of petroleum hydrocarbons in littoral and benthic environments: The TSESIS oil spill

The fate of saturated and aromatic hydrocarbons discharged into the coastal Baltic Sea environment from the TSESIS oil spill has been studied in the acute and postacute (one year) phases of the spill. Periodic samples of Mytilus edulis (mussels) from eight littoral zone stations and Macoma balthica from nine soft bottom stations were obtained as well as sediment trap samples and surface sediment samples. Glass capillary gas chromatography and gas chromatographic mass spectrometry were used as the analytical tools to determine saturated and aromatic hydrocarbon composition and concentrations in these samples.Sediment trap samples indicated that sizable quantities of chemically and microbially weathered oil were sedimented, and available for benthic uptake shortly after the spill. After initial uptake of sedimented oil (500 to 1000 μ/g dry weight), Macoma populations appear to have begun slow depuration through the first winter after the spill, but TSESIS oil was again introduced to the benthic stations studied during the following summer. Mytilus populations in the region were severely impacted by the oil. Initial depuration of spilled oil during the first month was rapid and nearly complete at all but the most heavily impacted stations one year after the spill. The post-spill depuration of assimilated hydrocarbons was characterised by a relative retention of alkylated dibenzothiophenes and alkylated phenanthrenes compared to their unsubstituted parent compounds and compared with the entire homologous naphthalene series.These data suggest that petroleum hydrocarbons from the TSESIS spill have become a chronic source of degraded saturated and aromatic hydrocarbons to the soft bottom benthic communities. Petroleum hydrocarbons in the benthic environment from this spill appear to reside in the difficulty sampled and mobile flucculent layer at the sediment/water interface and may affect epifaunal communities for an extended period of time.     

Numerical investigation of wave propagation in the Liverpool Bay, NW England

The computer model for near shore wave propagation,SWAN,was used to study wave climates in Liverpool Bay,northwest England with various input parameters,including bottom friction factor,white capping,wind drag formulation and effects of tidal modulations.Results were compared with in-situ measurements and reveal the impacts from these inputs on the predictions of wave height and propagation distributions.In particular,the model results were found very sensitive to different input formulations,and tend to underestimate the wave parameters under storm conditions in comparison with the observations.It is therefore important to further validate the model against detailed field measurements,particularly under large storms that are often of the primary concern.     

Pleistocene Ross seal (ommatophoca Rossi) from New Zealand (note)

A fossil seal jaw from the Early Pleistocene (Wanganui Series; Hautawan Stage) of Napier, New Zealand, is identified as that of an Ommatophoca rossi juvenile of less than 1.8 m body length.     

Institutional ambiguity in implementing the European Union Marine Strategy Framework Directive

This article addresses the institutional ambiguity that exists between the European, Regional and Member State levels in the implementation of the Marine Strategy Framework Directive (MSFD). The two main reasons for the emergence of institutional ambiguity are (1) the MSFD being a framework directive and (2) Member States are required to coordinate the implementation of the MFSD through the Regional Sea Conventions. Institutional ambiguity refers to the interference zone between different institutional settings that come together in new policy practices. New rules of the game are needed to bring these institutional settings together and the room to manoeuvre for the actors who negotiate these rules is a defining feature of institutional ambiguity. This article analyses the institutional ambiguity associated with MSFD implementation on the European and regional level for four European Seas: the North Sea, the Baltic Sea, the Mediterranean Sea and the Black Sea. The results indicate different levels of institutional ambiguity in each of the four regions, with the lowest level of ambiguity in the Baltic Sea and the highest in the Mediterranean Sea. Institutional ambiguity also exists on the European level, as coordination efforts have not resulted in clear directions for the implementation of the MSFD as yet. The level of institutional ambiguity is influenced by the relative number of EU member states bordering the particular sea and whether they consider implementation of the MFSD to be urgent. Member States bordering the Mediterranean and the Black Seas lack the support of Regional Sea Conventions in addition to receiving limited direction from the European level.