Numerical Effects (numerical + effects)

Distribution by Scientific Domains

Selected Abstracts

Nonnutrient anthropogenic chemicals in seagrass ecosystems: Fate and effects

Michael A. Lewis
Abstract Impacts of human-related chemicals, either alone or in combination with other stressors, are important to understand to prevent and reverse continuing worldwide seagrass declines. This review summarizes reported concentrations of anthropogenic chemicals in grass bed,associated surface waters, sediments, and plant tissues and phytotoxic concentrations. Fate information in seagrass-rooted sediments and overlying water is most available for trace metals. Toxicity results in aqueous exposures are available for at least 13 species and a variety of trace metals, pesticides, and petrochemicals. In contrast, results for chemical mixtures and chemicals in sediment matrices are uncommon. Contaminant bioaccumulation information is available for at least 23 species. The effects of plant age, tissue type, and time of collection have been commonly reported but not biological significance of the chemical residues. Experimental conditions have varied considerably in seagrass contaminant research and interspecific differences in chemical residues and chemical tolerances are common, which limits generalizations and extrapolations among species and chemicals. The few reported risk assessments have been usually local and limited to a few single chemicals and species representative of the south Australian and Mediterranean floras. Media-specific information describing exposure concentrations, toxic effect levels, and critical body burdens of common near-shore contaminants is needed for most species to support integrated risk assessments at multiple geographical scales and to evaluate the ability of numerical effects-based criteria to protect these marine angiosperms at risk. [source]

Reducing numerical diffusion in interfacial gravity wave simulations

O. B. Fringer
Abstract We demonstrate how the background potential energy is an excellent measure of the effective numerical diffusion or antidiffusion of an advection scheme by applying several advection schemes to a standing interfacial gravity wave. All existing advection schemes do not maintain the background potential energy because they are either diffusive, antidiffusive, or oscillatory. By taking advantage of the compressive nature of some schemes, which causes a decrease in the background potential energy, and the diffusive nature of others, which causes an increase in the background potential energy, we develop two background potential energy preserving advection schemes that are well-suited to study interfacial gravity waves at a density interface between two miscible fluids in closed domains such as lakes. The schemes employ total variation diminishing limiters and universal limiters in which the limiter is a function of both the upwind and local gradients as well as the background potential energy. The effectiveness of the schemes is validated by computing a sloshing interfacial gravity wave with a nonstaggered-grid Boussinesq solver, in which QUICK is employed for momentum and the pressure correction method is used, which is second-order accurate in time. For scalar advection, the present background potential energy preserving schemes are employed and compared to other TVD and non-TVD schemes, and we demonstrate that the schemes can control the change in the background potential energy due to numerical effects. Copyright 2005 John Wiley & Sons, Ltd. [source]

The evolution of substructure in galaxy, group and cluster haloes , III.

Comparison with simulations
ABSTRACT In a previous paper, we described a new method for including detailed information about substructure in semi-analytic models of halo formation based on merger trees. In this paper, we compare the predictions of our model with results from self-consistent numerical simulations. We find that in general the two methods agree extremely well, particularly once numerical effects and selection effects in the choice of haloes are taken into account. As expected from the original analyses of the simulations, we see some evidence for artificial overmerging in the innermost regions of the simulated haloes, either because substructure is being disrupted artificially or because the group-finding algorithms used to identify substructure are not detecting all the bound clumps in the highest-density regions. Our analytic results suggest that greater mass and force resolution may be required before numerical overmerging becomes negligible in all current applications. We discuss the implications of this result for observational and experimental tests of halo substructure, such as the analysis of discrepant magnification ratios in strongly lensed systems, terrestrial experiments to detect dark matter particles directly or indirect detection experiments searching for positrons, gamma-rays, neutrinos or other dark matter decay or annihilation products. [source]

Entropy sources in a dynamical core atmosphere model

Tim Woollings
Abstract Numerical atmosphere models are not generally constructed to ensure accurate treatment of entropy, but little is known about the significance of the resulting errors. This paper examines the entropy changes during a baroclinic wave simulation in a typical dynamical core model, specifically a ,-coordinate spectral model, which includes scale-selective dissipation terms in the form of a numerical hyperdiffusion. Lagrangian entropy conservation is found to be badly represented, with numerical transport errors resulting in cross-isentrope mass fluxes which are of the same size as those associated with some real diabatic processes. In a global average, the total entropy increases at a rate of just 0.5 mW m,2K,1. This, however, is seen to be the residual of two opposing numerical effects which are several times larger, namely the destruction of entropy by dispersion and Gibbs errors, and its creation by diffusion. The entropy generated by diffusion is shown to be remarkably insensitive to the details of the diffusion scheme. This leads us to hypothesize that the entropy source from diffusion is determined by the rate at which small scales are generated by the deformation field of the large-scale flow so that, while the diffusion mechanism is clearly unrealistic, the magnitude of the entropy source is, we argue, representative of that generated by physical dissipative processes in the real atmosphere. Even in this simple model it is not possible to quantify precisely the different entropy sources and sinks which combine to give the overall entropy change. However, we can say that if there is a systematic spurious entropy source in this model, then it is small, i.e. of size 0.5 mW m,2K,1 or smaller. Copyright 2006 Royal Meteorological Society [source]

Shear-driven magnetic buoyancy oscillations

V. Vermersch
Abstract The effects of uniform horizontal shear on a stably stratified layer of gas is studied. The system is initially destabilized by a magnetically buoyant flux tube pointing in the cross-stream direction. The shear amplifies the initial field to Lundquist numbers of about 200,400, but then its value drops to about 100,300, depending on the value of the sub-adiabatic gradient. The larger values correspond to cases where the stratification is strongly stable and nearly isothermal. At the end of the runs the magnetic field is nearly axisymmetric, i.e. uniform in the streamwise direction. In view of Cowling's theorem the sustainment of the field remains a puzzle and may be due to subtle numerical effects that have not yet been identified in detail. In the final state the strength of the magnetic field decreases with height in such a way that the field is expected to be unstable. Low amplitude oscillations are seen in the vertical velocity even at late times, suggesting that they might be persistent ( 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim) [source]