Maxime Beauchamp

A necessary distinction between spatial representativeness of an air quality monitoring station and the delimitation of exceedance areas

The European legislation on ambient air quality introduces the concepts of spatial representativeness of a monitoring station and spatial extent of an exceedance zone. Spatial representativeness is an essential macro-scale siting criterion which should be evaluated before the setting-up and during the life of a monitoring point. As for the exceedance area, it has to be defined each time an environmental objective is exceeded in an assessment zone. No specific approach is prescribed to delimit such areas. A probabilistic methodology is presented, based on a preliminary kriging estimation of atmospheric concentrations at each point of the domain. It is applied to NO2 pollution on the urban scale. In the proposed approach, a point belongs to the area of representativeness of a station if its concentration differs from the station measurement by less than a given threshold. To take the estimation uncertainty into account, the standard deviation of the kriging error is used in a probabilistic framework. The choice of the criteria used to deal with overlapping areas is first tested on NO2 annual mean concentration maps of France, built by combining surface monitoring observations and outputs from the CHIMERE chemistry transport model. At the local scale, data from passive sampling surveys and high -resolution auxiliary variables are used to provide a more precise estimation of the background pollution in different French cities. The traffic-related pollution can also be accounted for in the map by additional predictors such as distance to the road, and traffic-related NOx emissions. Similarly, the proposed approach is implemented to identify the points, at a given statistical risk, where the NO2 concentration is above the annual limit value.

A polynomial approximation of the traffic contributions for kriging-based interpolation of urban air quality model

Abstract The European directives for ambient air quality require to assess areas where air pollutant concentrations exceed a regulatory threshold. As the spatial distribution of the pollutant is not exactly known, deterministic atmospheric dispersion models are commonly used to supplement the observational network. To reduce the computational time, the simulations are made on irregular grids, especially in urban areas where the grid is refined close to the roads. An interpolation method is then necessary to map the dispersion model at any location. We propose a new geostatistical approach based on kriging with external drift to distinguish the information along and across the roads. An exponential function is introduced to describe the decrease of the concentrations across the roads. Its series expansion is used to build a set of polynomial auxiliary predictors with unknown coefficients. This framework leads to a drift that is more generic and flexible in the kriging system.