Catálogo de publicaciones - libros

Splitting techniques are, to our knowledge, used in all operationally run large scale air pollution models with many scenarios. The modellers believe that really huge computational tasks can be made tractable on the available computers by dividing them into a sequence of smaller and simple sub-tasks. Probably, the first simple splitting procedure for partial differential equations was proposed by Bagrinovski and Godunov in 1957. Since then many different splitting schemes were proposed and studied. A significant progress in splitting analysis was done during the last 3–4 years when the Laplace transformations technique was replaced by more powerful techniques. Several splitting schemes for large scale air pollution models are analysed and tested in a set of numerical experiments with respect to the splitting error. The computational properties of splitting schemes are analysed from algorithmic point of view. Parallel properties of these splitting schemes are also discussed.

Radon emissions from a radioactive waste disposal may constitute a major source of environment contamination and consequently a potential health hazard to the nearby population. Gaseous Radon-222 is generated from the radioactive decay of Radium-226 present in the tails. When it is formed, radon is free to diffuse along the pores of the residues to the surface and escape to the atmosphere.

Waste management and long term stabilisation has a major concern in reducing radon emissions to near-background levels. The common theoretical approach is done by calculating the cover thickness that allows a radon flux inferior to a stipulated and accepted value. The fundamentals of the conceptual model are based in the principles of diffusion across a porous medium, which allows the mathematical description of the radon transport through the waste and the cover. The basic diffusion equations are used for estimating the theoretical values of the radon flux formed from the decay of the Radium-226 contained in the waste material. The algorithm incorporates the radon attenuation originated by an arbitrary cover system placed over the radioactive waste disposal.

Once the radon is released into the atmosphere, it is available for atmospheric transport by the wind. Radon atmospheric dispersion is modelled by a modified Gaussian plume equation, which estimates the average dispersion of radon released from a point source representative of one or several uniform area sources. The model considers the medium point release between all the areas contaminated. The dispersion can be simulated in different wind directions, with different wind velocities, as well as in the dominant wind direction.

Regional forecasting is of great importance because of its potential for preventing economical and social consequences of natural and man-caused atmospheric pollution. A new non-standard approach and computational method is offered in this paper for the efficient solution for forecasting regional meteorological processes. The method replaces the Cauchy problem in the atmospheric model by a boundary-value problem and introduces a specific interpolation technique that has a number of advantages for the method to be computationally efficient. The model and method have been tested by the Hydrometeorological Centre of Ukraine and successfully applied in regional short- and middle-term weather forecasting for regions of Ukraine.

Application of tropospheric chemical transport models (CTMs) to chemical weather forecast on the global and regional scale is growing. In this paper advances and problems of numerical forecasts with Eulerian regional models are treated. After reviewing the progress of computer performance since the appearance of first complex CTMs, selected models are briefly introduced in order to demonstrate the spectrum of forecast activities. Ozone and particulate matter are chosen as examples of air quality forecasts on continental (European) and smaller spatial scales. The question of forecast and model evaluation is discussed and the potential of chemical data assimilation for the upcoming fourth generation of air quality forecast models emphasized.

There are various aspects of the air pollution, its measurement, modeling, control and effects. They might be considered to be ones of the most effective means of informing and educating university students in modeling problems of the atmospheric environment and air quality protection. Four experiments which might be carried out are described: collection and identification of biogenic matter in air, determination of the effect of ozone on rubber under laboratory conditions and evaluation of ozone in ambient air using the effect of rubber cracking, lichens as indicators of air pollution, and how an aerosol diffuses and how its diffusion is affected by temperature inversion.

The present paper aims at giving a general idea of the functions of influence and their application to air pollution and environmental problems, introducing a rather general formulation of the problem; application of the functions of influence in pollution abatement strategies and model sensitivity studies. The functions of influence had been originally defined for the case of linear chemistry as solutions of an equation conjugated to the pollution transport one. The problem of functions of influence in case of small disturbances of model parameters is introduced in the present paper, which makes it possible to apply the approach of functions of influence for the case of nonlinear chemistry. Some applications of the functions of influence in pollution abatement strategies, model sensitivity studies and model parameters optimization are discussed.

The Danish Eulerian Hemispheric Model (DEHM) developed at the National Environmental Research Institute (NERI) in Denmark has been applied for numerous studies of air pollution. DEHM was originally developed with the main purpose of investigating the atmospheric transport of pollutants to the Arctic region, but is now also applied for more local scale environmental problems. The DEHM model has therefore recently been developed with nesting capabilities, which allows for a large mother domain (the Northern Hemisphere) as well as one or two nested domains each with a three times higher horizontal resolution over a limited area. The chemistry version of the model (DEHM-REGINA) has recently been run for a period of 14 years with a nested domain covering Europe and driven by meteorological data from the nested MM5 model. A chemical scheme with 60 species is included and the long-term calculations were made possible due to the application of a new low-cost Linux cluster system recently implemented at NERI. Examples from the validation of the model show that the observed variations and tendencies on both daily to seasonal and longer time scales are well captured for both primary and secondary air pollution components.

A methodology of the use of the majorant (upper-percentile) concentration fields in the dispersion modelling for environmental security is discussed in this paper. General principles of constructing these fields are discussed, and an analytical expression for the majorant concentrations from the ground-level point source is presented here. A Russian regulatory dispersion model used for estimations of the air pollution from industrial sources had been formulated using these principles. Another Russian regulatory model dealing with accidental releases is described in this paper. This model is based on the dose approach and calculations of conditional majorants. It is used to determine the scale of contamination resulting from the accidental release of harmful chemical pollutants. Some results of validation of the models are also presented in this paper.

The implementation of rotated bilinear non-conforming finite elements in the advection-diffusion part of an Eulerian model for long-range transport of air pollutants is considered. The study is aimed at a possible implementation of such elements in the operational version of the Danish Eulerian Model (DEM).

One-dimensional linear finite elements are currently used during the space discretization of the advection-diffusion part of DEM. The application of more advanced numerical techniques in this part of the model is desirable.

Two variants of the rotated bilinear non-conforming finite element method (FEM) are analysed. They differ in the interpolation operator used to determine the local FEM basis of the related square reference element. In both cases, the nodal unknowns are associated with the midpoints of the edges.

The proposed monotone time-stepping scheme generalises the up-winding techniques originally developed for the case of finite difference method (FDM). It is of second order with respect to both space and time discretization parameters. The presented numerical results illustrate the behaviour of the considered new up-wind scheme. The well known in the community is used as a benchmark problem.

To control the emission of dangerous pollutants in order to ensure safety of human beings and ecosystems within prescribed limits becomes a common practice in the developed countries. The control policies are implemented as emission tax, limits on the surface traffic intensity, penalties, etc., and are usually targeted to a long-term effect. This paper presents and investigates an optimal control model of a short run (operative) control of the emissions of air pollutants in a bounded area, possibly exposed also to external pollution. The main scenario assumes that the safety bounds would be violated as a result of an abnormal additional pollution and/or unfavourable weather conditions (wind) if temporary restrictions on the emissions are not imposed. The aim of control is either to prevent violation of the safety bounds at a minimal cost (if possible at all), or to find an optimal trade-off between the cost of emission reduction and the damage caused by the pollution. Mathematically, the model involves a transport PDE with varying wind, (partly controllable) emission and non-linear deposition, and an objective function incorporating the cost of control and the cost of damage caused by the pollution. The theoretical background and a solution approach to the arising optimal control problem are presented in this paper. An analytic solution is obtained in a simplified scenario, which is useful for numerical tests.