Title: The development of methods to assess emissions at local scale

Primary Researcher: Alessandro Dosio
Email address: alessandro.dosio@wur.nl
Submitted on: July 17, 2003

Start date: 18 June 2001
End date: 17 June 2005

Introduction:

Current methods to determine emissions at local scales are often based on guidelines and handbooks in which emission-factors, based on emission measurements, are used. There is an urgent need to develop independent methods to determine emissions based on observations and dispersion models.
To be able to determine the optimal measuring strategy that should be used, state-of-the-art dispersion models and associated meteorological descriptions are required. Current models most in use are based on assumptions of steady state and homogeneous conditions. The real situation is often non-stationary and inhomogeneous, with complex terrain and variability in the spatial distribution of emissions and deposition. Mathematical models which solve explicitly the atmospheric turbulence (large eddy simulation) or which are based on the atmospheric surface layer similarity theory are in principle appropriate tools to determine the main features of dispersion processes on a local scale. For the simulation on local scales, the Lagrangian approach will be suited. This involves the modification/development of methods to model local turbulent exchange and to calculate concentrations. The coupling between calculated and measured concentrations will involve the use of inverse modelling and data-assimilation. The new methods will be tested against existing data sets of dispersion of emission sources at a local scale. Especially existing data-sets for N2O from grassland, for NH3 from farms and manure and for VOC from industrial complexes will be used for the evaluation. Also situations in urban areas, for example fine particle emissions from traffic, could be used. Based on the new developed and tested method, a measuring strategy for the determination of the emission-strength at local scales will be developed.

Aim:

The project is aimed at developing and testing a dedicated dispersion model in relation to a measuring strategy for determining emission rates from complex sources at local scale (order of magnitude 1 km).
The spatial and temporal evolution of pollutants in the atmosphere is commonly calculated using mathematical models. For regional to national scale and low resolution (5-10km) and national policy, the existing models (Nationaal Model and OPS) may do a good job. Nowadays there a tendency to more dedicated policy on a local scale. This implies, that local heterogeneity in sources, sinks and disturbances of the wind flow become more important in modelling. Meso scale models have the potential for better coping with complex terrain and higher resolution (Vonk et al,1999). These models are however not yet ready for application on a local scale.

Due to constraints of computer capacity, the model governing equations can only be solved with a limited temporal and spatial resolution. On space or time scales larger than the model resolution, the variables can be treated explicitly. The behavior of the variables on smaller scales, fluctuating or random component, has to be described by means of a parameterization.
Since the phenomena occur on smaller space (or time) scales than the grid size, these are commonly called subgrid scale (SGS) effects. The SGS effects are represented in the governing equations by the second-order moments of the velocity, temperature and concentration fields, namely fluxes and (co-)variances. One of the most important SGS is related to the non-uniform emission and deposition of the air pollutants.

Research:

Phase 1 (12 months)
Literature review.
Getting acquainted with air pollution dispersion and meteorology
Definition of data requirements for model use
Collect databases and make them suitable for comparison with model results.
Test the model for short and long-range transport (up to 10 km).
Compare the modelled results with results from statistical models.
Write a scientific paper on model comparison.

Phase 2 (15 months)
Determining the effect of space and time averaging by calculating concentration fluctuations, second-order moments.
Spectral analysis of the numerical experiments.
Sensitivity analysis for different configurations of the non-uniform emissions under various meteorological conditions.
Second scientific paper.

Future Research

Phase 3 (15 months) Sensitivity studies of the model grid resolution to obtain a good representation of the physical atmospheric processes (dispersion, turbulent mixing) that drive the distribution, and evolution of atmospheric species. Develop a measurement strategy to obtain source strength from diffuse sources with a minimum of measuring sites and measurements in height at a given level of uncertainty for the source strength estimate. Testing the developed strategy for low level diffuse sources for a near field and a far field situation. Third scientific paper in inverse modelling with the chosen model and developed measurement strategy.