PDF version-A +A


Contents of this page 

Program system for the spatial interpolation of point related data from monitoring stations. Applied for the State of Saxony. Brief sketch of procedure and results.


It was the aim of the development of IMMIKART, to create a software system for the spatial interpolation of point related data from the state monitoring stations of Saxony. Simple algorithms were supposed to be used, and the system should be operated later not by a scientist but by the staff members of the agency.

The air pollution concentrations in Saxony are monitored by 30 multi component stations, by different mobile stations and at 3 positions by the German Federal Environmental Agency - all stations are operated continuously. The results are point related concentrations for the positions of the stations.

As additional data input parameters were available the data from stations in the neighbouring states Saxony-Anhalt, Thuringia and Brandenburg and from the Czech Republic. Additionally, the digital terrain model, the digital land use model and the description of the monitoring stations was available.

The interpolation was first done for sulphur dioxide (SO2), suspended dust (SwSt), nitrogen dioxide (NO2) and ozone, for the annual mean and the 98-Perzentil, using a geostatistic - empirical procedure for a spatial resolution of 2.5 km x 2.5 km (IMMIKART part 1). Later it was enriched by benzene, soot and PM10, and a state wide dispersion calculation was done. Its results were combined with the results of the statistical-empirical methodology (IMMIKART part 2). See www.Lohmeyer.de/artikel/immikart2.htm.



The parameters, determining the concentrations at the different positions, were analysed at the beginning of the project and were made available for the spatial interpolation. This analysis showed whether the monitored concentrations are representative for the near- or for the far field. To determine this representativness of the concentration values measured at the individual measuring sites (as interpolation radiuses) among other informations the descrip­tions of the stations were used and the ratio of the local NO / NO2 concentration. Stations with very high NO / NO2 concentrations are strongly influenced from nearby nitrogen oxide sources, as for example heavily trafficked streets or agglomerations. Such stations have only a low spatial representativity if their data are used without accounting for that, whereas stations with low NO / NO2 - levels are representative for a big area surrounding them. The following multi-stage procedure leads to the requested concentration maps:

  1. The original monitoring data get corrections to account for altitude as well as proximity to strong air pollutant sources. Thus, also the results of monitoring stations, being strongly influenced by close air pollutant sources can be used for the spatial interpolation.
  2. On the basis of the data analysis, the monitoring stations are subdivided into 7 classes, for example city station, open country station etc.. To each of these classes a certain “interpolation radius” is assigned characterising the representativity of the corrected result.
  3. In a third step, a so called radial interpolation takes place to provide the basis concen­tration data, see Fig. 1.
  4. These basis concentrations were modified using secondary data as altitude, share of emission-relevant land use (settlement, city, highway) and share of forest. This was done in a different way for each pollutant according to the results of the data analysis, available results of concentration measurements and emission inventories for different Saxonian cities like Dresden, Chemnitz, Leipzig and Aue. Additionally empirically established relationships and information from the literature was used.

The results of the calculations are the spatially interpolated concentrations in Saxony for the year 1998. As an example for ozone see figure 2. For the production of such maps, standard software as for example ARCINFO or ARCVIEW can use the calculated data and do the graphical visualization.

The program system was set up in a way, enabling the LfUG of Saxony, to any arbitrary number of additional monitoring points. This showed to be necessary to include the results of special (say 1 year) monitoring programs or the results of dispersion calculations.



With IMMIKART part 2, a combination of the statistical - empirical procedure with results of dispersion calculations with the Lagrangian Dispersion Model LASAT is possible. This was done as the emission inventory for traffic in Saxony for the year 1999 became available. On the basis of this area covering emission inventory, dispersion calculations were executed resulting in the traffic-induced additional concentration for the whole of Saxony.

Compared to IMMIKART part 1 this procedure has the advantage of a higher spatially differentiated quality as well as the possibility of the preparation of predictions.

Additionally, a calculation module was developed, which determines the large scale back­ground concentration from the monitoring data of the regional monitoring system. The module adds these concentrations to the calculated traffic-induced additional concentration and balances these calculated concentrations with results of the field measurements. Both the annual means as well as the 98 percentiles were calculated, for PM10 additionally the 90 percentile.

Result of the calculations in the year 1998 are the spatially interpolated concentrations for benzene, soot and PM10 for whole area of Saxony. As example for PM10 see Fig. 3.

For questions please contact Mrs. Dipl.-Met. Antje Moldenhauer at info.dd(at)lohmeyer.de