George C. Efthimiou

COST 732 in practice: the MUST model evaluation exercise

The aim of this paper is to describe the use of a general methodology tailored to the evaluation of micro-scale meteorological models applied to flow and dispersion simulations in urban areas. This methodology, developed within COST 732, has been tested through a large modelling exercise involving many groups across Europe. The major test case used is the Mock Urban Setting Test (MUST) experiment representing an idealised urban area. It is emphasised that a full model evaluation is problem-dependent and requires several activities including a statistical validation that requires a careful choice of the metrics for the comparison with measurements.

Air dispersion modelling for individual exposure studies

The concentration fluctuations of a dispersing hazardous gaseous pollutant in the atmospheric boundary layer, and the hazard associated with short-term concentration levels, demonstrate the necessity of estimating the magnitude of these fluctuations using predicting models. To predict and estimate the maximum expected dosage and the exposure time within which the dosage exceeds certain health limits, the knowledge of the behaviour of concentration fluctuations at the point under consideration is needed. The whole effort is based on the field experiment MUST (Biltoft, 2001) and the computational simulations have been performed with the CFD code ADREA.

Atmospheric dispersion and individual exposure of hazardous materials. Validation and intercomparison studies

In a previous study, new approaches have been introduced in the CFD-RANS modelling, according to which the concentration time scales are estimated as a function not only of the flow turbulence time scales but also of the pollutant travel times. The new approaches have been implemented for the calculation of the concentration fluctuation dissipation time scale and the maximum individual exposure at short time intervals using the k-ζ model for turbulence parameterisation. The purpose of this study is to implement and validate again the new methodology using the widely known standard k-e model. The validation is performed using two selected trials of the MUST experiment under neutral conditions. Special emphasis is given on the selection of the constant value of the concentration fluctuation dissipation time scale when the k-e model is used. Also, an intercomparison of the results between the two turbulence models is performed with a view to identifying model strengths and limitations.

Statistical Projection of Material Intensity: Evidence from the Global Economy and 107 Countries

The material intensity (MI) of the economy remains among the most widely cited indicators in international statistics and reports, evaluating the efficient use and productivity of natural resources in the economic process. In the context of the contemporary economy‐wide material flow accounting framework, the material intensity of a country is evaluated through the estimation of the ratio of the domestic material consumption (DMC) to the gross domestic product (GDP) index (DMC/GDP). Indeed, the essential contribution of natural resources to the economic process requires the establishment of reliable projections of this intricate relationship to the future. These projections may provide critical information to policy makers and practitioners in order to evaluate the future dynamics of the efficient use of natural resources in the production process. Toward this objective, the present study evaluates and proposes an alternative novel methodology for MI statistical projections, based on the beta distribution, by using a deterministic model for predicting the maximum expected values. The parameters of the deterministic model are calculated from the estimated MI of the global economy. The evaluation of the model is then performed by using MI estimates from 107 individual countries. The agreement between the model and the estimates is very good. The proposed methods merit is its simplicity, as by using two statistics of the material intensity (mean and variance) and an integral time scale, it is feasible to calculate the probabilities of the MI of any country with a high degree of confidence.

A statistical model for the prediction of wind-speed probabilities in the atmospheric surface layer

Wind fields in the atmospheric surface layer (ASL) are highly three-dimensional and characterized by strong spatial and temporal variability. For various applications such as wind-comfort assessments and structural design, an understanding of potentially hazardous wind extremes is important. Statistical models are designed to facilitate conclusions about the occurrence probability of wind speeds based on the knowledge of low-order flow statistics. Being particularly interested in the upper tail regions we show that the statistical behaviour of near-surface wind speeds is adequately represented by the Beta distribution. By using the properties of the Beta probability density function in combination with a model for estimating extreme values based on readily available turbulence statistics, it is demonstrated that this novel modelling approach reliably predicts the upper margins of encountered wind speeds. The model’s basic parameter is derived from three substantially different calibrating datasets of flow in the ASL originating from boundary-layer wind-tunnel measurements and direct numerical simulation. Evaluating the model based on independent field observations of near-surface wind speeds shows a high level of agreement between the statistically modelled horizontal wind speeds and measurements. The results show that, based on knowledge of only a few simple flow statistics (mean wind speed, wind-speed fluctuations and integral time scales), the occurrence probability of velocity magnitudes at arbitrary flow locations in the ASL can be estimated with a high degree of confidence.

Source reconstruction of airborne toxics based on acute health effects information

The intentional or accidental release of airborne toxics poses great risk to the public health. During these incidents, the greatest factor of uncertainty is related to the location and rate of released substance, therefore, an information of high importance for emergency preparedness and response plans. A novel computational algorithm is proposed to estimate, efficiently, the location and release rate of an airborne toxic substance source based on health effects observations; data that can be readily available, in a real accident, contrary to actual measurements. The algorithm is demonstrated by deploying a semi-empirical dispersion model and Monte Carlo sampling on a simplified scenario. Input data are collected at varying receptor points for toxics concentrations (C; standard approach) and two new types: toxic load (TL) and health effects (HE; four levels). Estimated source characteristics are compared with scenario values. The use of TL required the least number of receptor points to estimate the release rate, and demonstrated the highest probability (>90%). HE required more receptor points, than C, but with lesser deviations while probability was comparable, if not better. Finally, the algorithm assessed very accurately the source location when using C and TL with comparable confidence, but HE demonstrated significantly lower confidence.

Modelling Short-Term Maximum Individual Exposure from Airborne Hazardous Releases in Urban Environments. Part ΙI: Validation of a Deterministic Model with Wind Tunnel Experimental Data

The capability to predict short-term maximum individual exposure is very important for several applications including, for example, deliberate/accidental release of hazardous substances, odour fluctuations or material flammability level exceedance. Recently, authors have proposed a simple approach relating maximum individual exposure to parameters such as the fluctuation intensity and the concentration integral time scale. In the first part of this study (Part I), the methodology was validated against field measurements, which are governed by the natural variability of atmospheric boundary conditions. In Part II of this study, an in-depth validation of the approach is performed using reference data recorded under truly stationary and well documented flow conditions. For this reason, a boundary-layer wind-tunnel experiment was used. The experimental dataset includes 196 time-resolved concentration measurements which detect the dispersion from a continuous point source within an urban model of semi-idealized complexity. The data analysis allowed the improvement of an important model parameter. The model performed very well in predicting the maximum individual exposure, presenting a factor of two of observations equal to 95%. For large time intervals, an exponential correction term has been introduced in the model based on the experimental observations. The new model is capable of predicting all time intervals giving an overall factor of two of observations equal to 100%.

Validation of an Inverse Method for the Source Determination of a Hazardous Airborne Material Released from a Point Source in an Urban Environment

An improved inverse method was presented recently for the estimation of the location and the rate of an unknown point stationary source of passive atmospheric pollutant in a complex urban geometry. The inverse method was incorporated in the well-established and updated version of the ADREA-HF Computational Fluid Dynamics code. The key improvement of the proposed inverse method implementation lies in a two-step segregated approach combining a correlation and cost functions. At first only the source coordinates are analyzed using a correlation function of measured and calculated concentrations. In the second step the source rate is identified by minimizing a quadratic cost function. The validation of the new algorithm is performed by simulating the MUST wind tunnel experiment. Overall, we observed significant improvement, compared to previous implementations, on reconstructing the source information (location and rate).