Flexible self-adjustment of task deployment in dynamic environments

This paper presents an investigation into dynamic self-adjustment of task deployment and other aspects of self-management, through the embedding of multiple policies. Non-dedicated loosely-coupled computing environments, such as clusters and grids are increasingly popular platforms for parallel processing. These abundant systems are highly dynamic environments in which many sources of variability affect the run-time efficiency of tasks. The dynamism is exacerbated by the incorporation of mobile devices and wireless communication. This paper proposes an adaptive strategy for the flexible run-time deployment of tasks; to continuously maintain efficiency despite the environmental variability. The strategy centres on policy-based scheduling which is informed by contextual and environmental inputs such as variance in the round-trip communication time between a client and its workers and the effective processing performance of each worker. A self-management framework has been implemented for evaluation purposes. The framework integrates several policy-controlled, adaptive services with the application code, enabling the run-time behaviour to be adapted to contextual and environmental conditions. Using this framework, an exemplar self-managing parallel application is implemented and used to investigate the extent of the benefits of the strategy

Policy-based techniques for self-managing parallel applications

This paper presents an empirical investigation of policy-based self-management techniques for parallel applications executing in loosely-coupled environments. The dynamic and heterogeneous nature of these environments is discussed and the special considerations for parallel applications are identified. An adaptive strategy for the run-time deployment of tasks of parallel applications is presented. The strategy is based on embedding numerous policies which are informed by contextual and environmental inputs. The policies govern various aspects of behaviour, enhancing flexibility so that the goals of efficiency and performance are achieved despite high levels of environmental variability. A prototype self-managing parallel application is used as a vehicle to explore the feasibility and benefits of the strategy. In particular, several aspects of stability are investigated. The implementation and behaviour of three policies are discussed and sample results examined.

Reliability based design optimisation for system-in-package

This paper discusses a reliability based optimisation modelling approach demonstrated for the design of a SiP structure integrated by stacking dies one upon the other. In this investigation the focus is on the strategy for handling the uncertainties in the package design inputs and their implementation into the design optimisation modelling framework. The analysis of fhermo-mechanical behaviour of the package is utilised to predict the fatigue life-time of the lead-free board level solder interconnects and warpage of the package under thermal cycling. The SiP characterisation is obtained through the exploitation of Reduced Order Models (ROM) constructed using high fidelity analysis and Design of Experiments (DoE) methods. The design task is to identify the optimal SiP design specification by varying several package input parameters so that a specified target reliability of the solder joints is achieved and in the same time design requirements and package performance criteria are met

Personalised e-learning, semantic web and learning ontologies

Kurzel(2004) points out that researchers in e-learning and educational technologists, in a quest to provide improved Learning Environments (LE) for students are focusing on personalising the experience through a Learning Management System (LMS) that attempts to tailor the LE to the individual (see amongst others Eklund & Brusilovsky, 1998; Kurzel, Slay, & Hagenus, 2003; Martinez,2000; Sampson, Karagiannidis, & Kinshuk, 2002; Voigt & Swatman; 2003). According to Kurzel (2004) this tailoring can have an impact on content and how it’s accessed; the media forms used; method of instruction employed and the learning styles supported. This project is aiming to move personalisation forward to the next generation, by tackling the issue of Personalised e-Learning platforms as pre-requisites for building and generating individualised learning solutions. The proposed development is to create an e-learning platform with personalisation built-in. This personalisation is proposed to be set from different levels of within the system starting from being guided by the information that the user inputs into the system down to the lower level of being set using information inferred by the system’s processing engine. This paper will discuss some of our early work and ideas.

Semantic web and e-learning

With emergence of "Semantic Web" there has been much discussion about the impact of technologies such as XML and RDF on the way we use the Web for developing e-learning applications and perhaps more importantly on how we can personalise these applications. Personalisation of e-learning is viewed by many authors (see amongst others Eklund & Brusilovsky, 1998; Kurzel, Slay, & Hagenus, 2003; Martinez, 2000; Sampson, Karagiannidis, & Kinshuk, 2002; Voigt & Swatman, 2003) as the key challenge for the learning technologists. According to Kurzel (2004) the tailoring of e-learning applications can have an impact on content and how it's accesses; the media forms used; method of instruction employed and the learning styles supported. This paper will report on a research project currently underway at the eCentre in University of Greenwich which is exploring different approaches and methodologies to create an e-learning platform with personalisation built-in. This personalisation is proposed to be set from different levels of within the system starting from being guided by the information that the user inputs into the system down to the lower level of being set using information inferred by the system's processing engine.

Modelling dosator filling and discharge of powder

Dosators and other dosing mechanisms operating on generally similar principles are very widely used in the pharmaceutical industry for capsule filling, and for dosing products that are delivered to the customer in powder form such as inhalers. This is a trend that is set to increase. However a significant problem for this technology is being able to predict how accurately and reliably, new drug formulations will be dosed from these machines prior to manufacture. This paper presents a review of the literature relating to powder dosators which considers mathematical models for predicting dosator performance, the effects of the dosator geometry and machine settings on the accuracy of the dose weight. An overview of a model based on classical powder mechanics theory that has been developed at The University of Greenwich is presented. The model uses inputs from a range of powder characterisation tests including, wall friction, bulk density, stress ratio and permeability. To validate the model it is anticipated that it will be trialled for a range of powders alongside a single shot dosator test rig.

Heavy metals in the Fal estuary, Cornwall: A study of long term contamination by mining waste and its effects on estuarine organisms. Occasional Publication of the Marine Biological Association 2

The Fal Estuary System in West Cornwall has, over many centuries, received inputs of heavy metals from various mining activities. In this context its most important tributary is the Carnon River. Analyses of organisms from the Fal Estuary have shown that some species contain abnormally high concentrations of Cu, Zn and As, especially those living in Restronguet Creek.

Particulate and dissolved aluminum and titanium in the upper water column of the Atlantic Ocean

This study investigates the oceanic behavior of the lithogenic trace elements Al and Ti in the upper 200 m of the Atlantic Ocean. The distribution of both metals in the dissolved and particulate phases was assessed along an E-W transect in the eastern tropical North Atlantic (December 2009) and along a meridional Atlantic transect (April-May 2010). The surface water concentrations of particulate and dissolved Al and Ti reflected the previously observed pattern of atmospheric inputs into the Atlantic Ocean. Subsurface minima at stations with pronounced fluorescence maxima were observed, suggesting a link between biological productivity and the removal of both dissolved and particulate Al and Ti. This may include uptake mechanisms, adsorption and aggregation processes on biogenic particle surfaces and the formation of large, fast sinking biogenic particles, e.g., fecal pellets. Residence times in the upper water column (100 m) of the tropical and subtropical North Atlantic were estimated to range in the order of days to weeks in the particulate phases (Al: 3-22 days, Ti: 4-37 days) and were 0.9-3.8 years for Al and 10-31 years for Ti in the dissolved phases. Longer residence times in both phases in the South Atlantic are consistent with lower biological productivity and decreased removal rates. In the upper water column, Al was predominantly present in the dissolved form, whereas Ti mostly occurred in the particulate form. Largest deviations in the partition coefficients between the particulate and dissolved phases were found in the surface waters, together with excess dissolved Al over Ti compared to the crustal source. This likely reflects elevated dissolution of Al compared to Ti from atmospheric mineral particles.

Unbiased Linear Property Estimation For Spheres From Sections Exhibiting Overprojection And Truncation

In the biological sciences, stereological techniques are frequently used to infer changes in structural parameters (volume fraction, for example) between samples from different populations or subject to differing treatment regimes. Non-homogeneity of these parameters is virtually guaranteed, both between experimental animals and within the organ under consideration. A two-stage strategy is then desirable, the first stage involving unbiased estimation of the required parameter, separately for each experimental unit, the latter being defined as a subset of the organ for which homogeneity can reasonably be assumed. In the second stage, these point estimates are used as data inputs to a hierarchical analysis of variance, to distinguish treatment effects from variability between animals, for example. Techniques are therefore required for unbiased estimation of parameters from potentially small numbers of sample profiles. This paper derives unbiased estimates of linear properties in one special case—the sampling of spherical particles by transmission microscopy, when the section thickness is not negligible and the resulting circular profiles are subject to lower truncation. The derivation uses the general integral equation formulation of Nicholson (1970); the resulting formulae are simplified, algebraically, and their efficient computation discussed. Bias arising from variability in slice thickness is shown to be negligible in typical cases. The strategy is illustrated for data examining the effects, on the secondary lysosomes in the digestive cells, of exposure of the common mussel to hydrocarbons. Prolonged exposure, at 30 μg 1−1 total oil-derived hydrocarbons, is seen to increase the average volume of a lysosome, and the volume fraction that lysosomes occupy, but to reduce their number.

Aquatic Distribution And Heterotrophic Degradation Of Polycyclic Aromatic-Hydrocarbons (PAH) In The Tamar Estuary

Variations in the concentrations and microheterotrophic degradation rates of selected Polycyclic Aromatic Hydrocarbons (PAH) in the water column of the Tamar Estuary were investigated in relation to the major environmental variables. Concentrations of individual PAH varied typically between i and 50 ng l−1 Based on their observed environmental behaviour the PAH appeared divisible into two groupings: (1) low molecular weight PAH incorporating naphthalene, phenanthrene and anthracence and (a) the larger molecular weight homologues (fluoranthene, pyrene, chrysene, benz(a)anthracene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)-pyrene). Group 1 PAH showed a complex distribution throughout the estuary with no significant correlations with either salinity or suspended particulates. Based on their relatively low particle affinity and high water solubilities and vapour pressures, volatilization is proposed as an important process in determining their fate. Microheterotrophic turnover times of naphthalene varied between x and 30 days, and were independent of suspended solids with maximum degradation rates located in the central and urban regions of the Estuary. When compared with the flushing times for the Tamar (3–5 days), it is probable that heterotrophic activity is important in the removal of naphthalene (and possibly the other Group 1 PAH) from the estuarine environment. In contrast Group 2 PAH concentrations exhibited highly significant correlations with suspended particulates. Highest concentrations occurred at the turbidity maximum, with a secondary concentration maximum localized to the industrialized portion of the estuary and associated with anthropogenic inputs. Laboratory degradation studies of benzo(a)pyrene in water samples taken from the estuary showed turnover times for the compound of between 2000 and 9000 days. Degradation rates correlated positively with suspended solids. The high particulate affinity and microbial refractivity of Group 2 PAH indicate sediment burial as the principal tate of these PAH in the Tamar Estuary. Estuarine sediments contained typically 50–1500 ng g−1 dry weight of individual PAH which were comparable to the levels of Group 2 PAH associated with the suspended particulates. Highest concentrations occurred at the riverine end of the estuary resulting from unresolved inputs in the catchment. Subsequent dilution by less polluted marine sediments together with slow degradation results in a seaward trend of decreasing concentrations. However, there is a secondary maximum of PAH superimposed on this trend which is associated with urban Plymouth.

Simulating The Impact Of Technological-Change On Dissolved Cadmium Distribution In The Severn Estuary

An estuarine model is described which computes the dispersive and advective properties of the Severn Estuary. It was calibrated and validated using 50 measured salinity distributions and then used to predict the magnitude and sitings of the major inputs of dissolved cadmium levels throughout the estuary. The results provided an impetus for implementing tighter controls on effluents and for improving estimates of cadmium discharges from industrial sources. The model has also been used to investigate the sensitivity of the estuarine system to changes in dispersion; by considering large reductions in the dispersion coefficients it is hoped that the results might be indicative of the environmental consequences following the construction of a tidal power generating scheme.

DISCOVERY 2010: Spatial and temporal variability in a dynamic polar ecosystem

The Scotia Sea has been a focus of biological- and physical oceanographic study since the Discovery expeditions in the early 1900s. It is a physically energetic region with some of the highest levels of productivity in the Southern Ocean. It is also a region within which there have been greater than average levels of change in upper water column temperature. We describe the results of three cruises transecting the central Scotia Sea from south to north in consecutive years and covering spring, summer and autumn periods. We also report on some community level syntheses using both current-day and historical data from this region. A wide range of parameters were measured during the field campaigns, covering the physical oceanography of the region, air–sea CO2 fluxes, macro- and micronutrient concentrations, the composition and biomass of the nano-, micro- and mesoplankton communities, and the distribution and biomass of Antarctic krill and mesopelagic fish. Process studies examined the effect of iron-stress on the physiology of primary producers, reproduction and egestion in Antarctic krill and the transfer of stable isotopes between trophic layers, from primary consumers up to birds and seals. Community level syntheses included an examination of the biomass-spectra, food-web modelling, spatial analysis of multiple trophic layers and historical species distributions. The spatial analyses in particular identified two distinct community types: a northern warmer water community and a southern cold community, their boundary being broadly consistent with the position of the Southern Antarctic Circumpolar Current Front (SACCF). Temperature and ice cover appeared to be the dominant, over-riding factors in driving this pattern. Extensive phytoplankton blooms were a major feature of the surveys, and were persistent in areas such as South Georgia. In situ and bioassay measurements emphasised the important role of iron inputs as facilitators of these blooms. Based on seasonal DIC deficits, the South Georgia bloom was found to contain the strongest seasonal carbon uptake in the ice-free zone of the Southern Ocean. The surveys also encountered low-production, iron-limited regions, a situation more typical of the wider Southern Ocean. The response of primary and secondary consumers to spatial and temporal heterogeneity in production was complex. Many of the life-cycles of small pelagic organisms showed a close coupling to the seasonal cycle of food availability. For instance, Antarctic krill showed a dependence on early, non-ice-associated blooms to facilitate early reproduction. Strategies to buffer against environmental variability were also examined, such as the prevalence of multiyear life-cycles and variability in energy storage levels. Such traits were seen to influence the way in which Scotia Sea communities were structured, with biomass levels in the larger size classes being higher than in other ocean regions. Seasonal development also altered trophic function, with the trophic level of higher predators increasing through the course of the year as additional predator-prey interactions emerged in the lower trophic levels. Finally, our studies re-emphasised the role that the simple phytoplankton-krill-higher predator food chain plays in this Southern Ocean region, particularly south of the SACCF. To the north, alternative food chains, such as those involving copepods, macrozooplankton and mesopelagic fish, were increasingly important. Continued ocean warming in this region is likely to increase the prevalence of such alternative such food chains with Antarctic krill predicted to move southwards.

Low colloidal associations of aluminium and titanium in surface waters of the tropical Atlantic

The distribution of dissolved, soluble and colloidal fractions of Al and Ti was assessed by ultrafiltration studies in the upper water column of the eastern tropical North Atlantic. The dissolved fractions of both metals were found to be dominated by the soluble phase smaller than 10 kDa. The colloidal associations were very low (0.2–3.4%) for Al and not detectable for Ti. These findings are in some contrast to previous estimations for Ti and to the predominant occurrence of both metals as hydrolyzed species in seawater. However, low tendencies to form inorganic colloids can be expected, as in seawater dissolved Al and dissolved Ti are present within their inorganic solubility levels. In addition, association with functional organic groups in the colloidal phase is unlikely for both metals. Vertical distributions of the dissolved fractions showed surface maxima with up to 43 nM of Al and 157 pM of Ti, reflecting their predominant supply from atmospheric sources to the open ocean. In the surface waters, excess dissolved Al over dissolved Ti was present compared to the crustal source, indicating higher solubility and thus elevated inputs of dissolved Al from atmospheric mineral particles. At most stations, subsurface minima of Al and Ti were observed and can be ascribed to scavenging processes and/or biological uptake. The dissolved Al concentrations decreased by 80–90% from the surface maximum to the subsurface minimum. Estimated residence times in the upper 100 m of the water column ranged between 1.6 and 4 years for dissolved Al and between 14 and 17 years for dissolved Ti. The short residence times are in some contrast to the low colloidal associations of Al and Ti and the assumed role of colloids as intermediates in scavenging processes. This suggests that either the removal of both metals occurs predominantly via direct transfer of the hydrolyzed species into the particulate fraction or that the colloidal phase is rapidly turned over in the upper water column.

The carbonate system of the NW European shelf: sensitivity and model validation

The ocean plays an important role in regulating the climate, acting as a sink for carbon dioxide, perturbing the carbonate system and resulting in a slow decrease of seawater pH. Understanding the dynamics of the carbonate system in shelf sea regions is necessary to evaluate the impact of Ocean Acidification (OA) in these societally important ecosystems. Complex hydrodynamic and ecosystem coupled models provide a method of capturing the significant heterogeneity of these areas. However rigorous validation is essential to properly assess the reliability of such models. The coupled model POLCOMS–ERSEM has been implemented in the North Western European shelf with a new parameterization for alkalinity explicitly accounting for riverine inputs and the influence of biological processes. The model has been validated in a like with like comparison with North Sea data from the CANOBA dataset. The model shows good to reasonable agreement for the principal variables, physical (temperature and salinity), biogeochemical (nutrients) and carbonate system (dissolved inorganic carbon and total alkalinity), but simulation of the derived variables, pH and pCO2, are not yet fully satisfactory. This high uncertainty is attributed mostly to riverine forcing and primary production. This study suggests that the model is a useful tool to provide information on Ocean Acidification scenarios, but uncertainty on pH and pCO2 needs to be reduced, particularly when impacts of OA on ecosystem functions are included in the model systems.

Nutrient budgets for European seas: A measure of the effectiveness of nutrient reduction policies

Socio-economic development in Europe has exerted increasing pressure on the marine environment. Eutrophication, caused by nutrient enrichment, is evident in regions of all European seas. Its severity varies but has, in places, adversely impacted socio-economic activities. This paper aims to evaluate the effectiveness of recently adopted policies to reduce anthropogenic nutrient inputs to European seas. Nitrogen and phosphorus budgets were constructed for three different periods (prior to severe eutrophication, during severe eutrophication and contemporary) to capture changes in the relative importance of different nutrient sources in four European seas suffering from eutrophication (Baltic Proper, coastal North Sea, Northern Adriatic and North-Western Black Sea Shelf). Policy success is evident for point sources, notably for P in the Baltic and North Seas, but reduction of diffuse sources has been more problematic.