Joseph Alcamo

A Framework for Error Analysis of a Long-Range Transport Model with Emphasis on Parameter Uncertainty

A comprehensive framework for model error analysis is applied to the EMEP-W model of longrange transport of sulfur in Europe. This framework includes a proposed taxonomy of model uncertainties. Parameter uncertainties were investigated by Monte Carlo simulation of two source-receptor combinations. A 20% input parameter uncertainty (expressed as a coefficient of variation = standard deviation/mean) yielded a 15–22% output error of total sulfur deposition. The relationship between output error and input uncertainty was approximately proportional. Covariance between parameters can have an important effect on computed model error, and can either exaggerate or reduce errors compared to the uncorrelated case. Of the model state variables, SO2 air concentration and wet deposition had the highest error, and total sulfur deposition the lowest. It was also found that it is more important to specify the dispersion of the input parameter frequency distributions than their shape. The results of the model error analysis were applied to routine calculations of deposition in Europe. An error (coefficient of variation) of 20% for transfer coefficients throughout Europe yielded spatial variations in the order of a few tens to a few hundreds of km in computed deposition isolines of 2 and 5 g sulfur m−2a−1.

The uncertainty of atmospheric source-receptor relationships in Europe

Abstract This paper investigates the composite uncertainty of a long range transport model of sulfur in Europe. This composite uncertainty includes the effect on model output of uncertain transport wind, meteorological forcing functions, parameters, and spatial distribution of emissions. Other sources of uncertainty are omitted in the analysis. Stochastic simulation is used for computations. The combined effect of these uncertainties on total annual sulfur deposition at three receptors ranged from about 10 to 20% (coefficient of variation). In comparing the effects of different uncertainties on annual model output, meteorological forcing functions were found to be the least important because of their frequent temporal variation in the model. The stochastic procedure was also used to compute the uncertainty of transfer coefficients for 30 source-receptor combinations; their relative uncertainty ranged from about 10 to 30% and was not correlated with distance. However, their absolute uncertainty (standard deviation) was strongly correlated. with distance and was found to be proportional to the values of the transfer coefficients themselves. This insight was used to develop a simple method for estimating the uncertainty of sulfur deposition calculated with a transfer matrix. This method was then used to evaluated the “reliability” of emission reduction scenarios in reducing deposition.

Computing heavy metals in Europe's atmosphere—I. Model development and testing

Abstract This paper presents the development of the TRACE model (TRace toxic Air Concentrations in Europe) which computers the air concentration and deposition of various heavy metals (As, Cd, Pb, Zn) on a European scale. TRACE is an improved climatological-type model in that (1) travel time is computed from an empirical function rather than from an assumed constant velocity, (2) the model tends to conserve mass, (3) the irregularity of spatial deposition patterns is captured, and (4) parameters are objectively determined. The dry deposition velocity is spatially varying, and is computed with a dry deposition model as a function of “local” u ∗ , z o , together with an assumed characteristic particle size distribution. The model has been used to compute levels of heavy metals for 1978–1985 throughout Europe. Calculations agree with As and Pb observations with a factor of two, and underestimate Cd and Zn observations. Using the model it was estimated that wet deposition exceeds dry deposition in most of central Europe. The mean residence time of the mass of As, Cd, Pb and Zn in Europes lower atmosphere is estimated to be about 64h and for Pb, 96 h.

Modeling Heavy Metals in Europe’s Atmosphere: A Combined Trajectory-Climatologic Approach

The long range transport of heavy metals in the atmosphere leads to low but steady deposition of heavy metals into soils, lakes and forests in Europe. Although their atmospheric fluxes are low, these metals can accumulate in the detritus of soils. This accumulated metal may be mobilized by acidification and may then disturb soil organisms, which in turn can lead to disturbance of organic matter decomposition and nutrient cycling (Ottar et al., 1989). Circumstantial evidence from wind sector analysis and trajectory modeling (Pacyna et al., 1984), and receptor modeling (Stevens et al., 1984) indicates that the deposition of metals may originate from sources many hundreds of kilometers away. It is thought that large particles (> 10 μ) settle out in the vicinity of the stacks but stack gases containing metals condense into small particles which then coalesce within hours into particles with diameters of 0.1 to 1.0 μ; particles of this size are too coarse to be efficiently removed by brownian diffusion yet too small to settle out by gravity. Hence these metal particles may travel long distances before being removed by precipitation or dry deposition.

The effects of selected NOx reduction scenarios on long-term nitrogen deposition and episodic ozone levels in Europe

For computational reasons, evaluations of NO(x) emission controls usually concentrate on either episodic or annual impacts on pollution or deposition levels. However, previously published model results indicate that the consequences of NO(x) controls can be quite different on these different time scales. In this paper we analyse the impact of a consistent set of NO(x) control scenarios on both the episodic and annual time-scales. Using similar models, we compute levels of episode peak O(3) and NO(2) and annual NO(y)-N and total N deposition at three locations in Europe due to six emission scenarios derived from OECD estimates. An NO(x) control scenario which reduces European emissions by 63%, only results in total annual N deposition reductions of 19, 36 and 26% at the three locations examined because of the influence of ammonia-nitrogen deposition. The same scenario results in either increases or decreases in episode peak O(3) due to the influence of hydrocarbons. Emission reduction strategies should take into account not only NO(x) emissions, but emissions of other pollutants, such as hydrocarbons and ammonia.

Water and fire: Water needs of future coal development in the Soviet Union and the United States

This paper presents estimates of water requirements for future coal use in the USSR and the U.S. Future levels of coal use were based on scenarios presented by IIASA in Energy in a Finite World. As a first step in the analysis, IIASAs coal scenarios were broken down from the scale of ‘world region’ to the scale of coal-producing region. This exercise revealed that American and Soviet coal targets, which seem feasible when viewed on the ‘world-region’ scale, may be difficult to attain on the coal-region scale due to insufficient coal reserves in some regions. n nIn the next stage of the analysis, an analytical model was developed, which describes on the coal-region scale the quantity of water required during different stages of coal development from mining to its final conversion to useful energy. Application of this model to each of ten principal coal-producing regions of the US and USSR suggested that roughly 1–2 tons of water will be consumed for every ton-equivalent (tce) of coal-fuel delivered. However, these estimates assume a high degree of water conservation; with less emphasis on conservation, perhaps 50% more water will be required. n nWater requirements for coal were then compared with competitive water uses in each U.S. coal region, as well as estimates of surface water supply in these regions. It was found that the amount of water needed for coal is small relative to other projected water uses such as agriculture and industry. However, after accounting for competitive water uses, there will probably be little or no water available for coal use during dry years in the Southwest and Northwest regions. Unless significant quantities of water can be stored for these years, coal development will have to displace other water uses in these regions. n nIntense water pressure will probably also occur in the Asian-USSR coal region of Ekibastuz, and possibly in Kuznetsk, Kansk-Achinsk, and Tungusska. n nIt is concluded, therefore, that an overall four- or fivefold expansion of coal use in the U.S. and Soviet Union will probably be constrained to some degree by both limited coal reserves and lack of readily available water.

A regional perspective of global coal scenarios

Energy planning from a global perspective sometimes overlooks constraints which occur on a smaller geographic level. This communication presents an example of constraints which arise when coal scenarios, based on global modelling, are applied to specific coal regions in the USA and USSR. It was found that coal production and use may be constrained in certain regions by limited coal reserves as well as insufficient surface water during drought years.

Model Simulation of the Atmospheric Input of Trace Metals into the North, Baltic, Mediterranean and Black Seas

Since the middle of this century, energy generation, industrial production and transportation have caused serious environmental contamination by trace elements including heavy metals. The rate of contamination can vary from place to place as a function of source densities and intensities of heavy metals flux as well as meteorological conditions. Aerosols containing heavy metals can be transported far away from their sources by advection before being deposited.

Results from a Climatological Model of Heavy Metals in Europe’s Atmosphere

There is increasing observational evidence that heavy metals’ air emissions cause not only local contamination, but also travel long distances in Europe and contribute to widespread, although low-level, contamination of the environment (e.g. Pacyna, et al, 1984; Ottar, et al, 1989). The TRACE model (TRB.cetoxic Air Concentrations in fftirope) has been developed to compute the long range transport of various heavy metals (As, Cd, Pb, and Zn) on the European-scale. A preliminary version of this model was reported in Alcamo, et al (1991a) and model refinements and application are given in Alcamo, et al (1991b) and Bartnicki et al (1991).

Atmospheric models and acidification: Summary and conclusions of the Warsaw II meeting on atmospheric computations to assess acidification in Europe

Three topics are discussed in this report: sensitivity/uncertainty analysis of long range transport models, the interface between atmospheric models of different scales, and linkage between atmospheric and ecological models.In separate analyses of long range transport models, it was found that uncertainty of annual S deposition was mostly affected by uncertainty of wind velocity, mixing height and wet deposition parameterization. Uncertain parameters collectively caused S deposition errors of around 10–25% (coefficient of variation) in the models examined. The effect of interannual meteorological variability on computed annual S deposition was relatively small.Different methods were presented for combining models of regional and interregional scale. It was found to be more important to include interregional information in regional-scale models for annual computations compared to episodic computations.A variety of linkage problems were noted between atmospheric and ecological models. The vertical distribution of pollutants and ‘forest fittering’ of pollutant deposition were found to be important in ecological impact calculations but lacking in the output of most interregional atmospheric models.