As a part of an Air Quality Management project in Tallinn, measurement campaigns in order to generate data with high spatial resolution have been performed. In Tallinn, databases containing high time resolution data from a few locations already exist, and these two measurement databases will together provide a good basis for future validation of dispersion model results.

Measurements of SO₂, NO₂ and VOC were carried out using diffusive samplers and were organised as a total of 12 monthly campaigns during the period from April 1996 to January 1998. The diffusive samplers give average air concentrations with continuous time coverage from a few days up to a month. The technique is based on the concept of molecular diffusion, where the gas molecules diffuse into a sampler where they are quantitatively collected. The samplers are easy to handle, and the method does not require electricity or calibration in the field.

In order to map concentration levels of the measured pollutants in different typical local environments within the urban area of Tallinn, a classification of sites was made according to exposure to road vehicle emissions and street ventilation conditions. Four different categories of sites, from busy street environment with poor ventilation to local background conditions representing a general pollution level were classified. In addition to these, measurements at background locations were also made. A total of between 20-80 sites of different classification were used simultaneously during the campaign measurements.

The measured monthly levels of SO₂ and NO₂ were, as expected, different at the differens site categories. In the table below, the results from all 12 measurement campaigns are summarized.

	NO2, monthly averages, µg/m3			SO2, monthly averages, µg/m3
	Max	Average	Min	Max	Average	Min
Busy street, poor ventilation	48	36	27	15	10	4
Busy street, good ventilation	50	29	10	21	8	1
Residential area	33	18	6	19	7	1
Local Background	30	14	5	17	6	1
Background	16	8	1	12	4	1

There were larger differences in concentration levels between the site categories for NO₂ than for SO₂. The different concentration patterns for the two components are probably mainly due to the emission situation. NO₂ is mostly generated by traffic, hence the highest concentrations will be found close to roads with high traffic intensity. The sources of SO₂ can be assumed to be more widespread within the area studied, since a major part of the SO₂ emissions originate from fossil fuel combustion in large power plants as well as in local residential heating devices.

There was a seasonality of higher concentration levels in wintertime than during summer of SO₂ at all site categories, and for NO₂ at the less exposed sites. At the two most exposed site categories no evident seasonality in NO₂ levels was detected.

The average levels of VOC, all measured at busy street environments with good ventilation, ranged from 19-95 µg/m³ of total VOC (eight components) at the different sites during four campaign months. The corresponding leves of benzene were between 3-13 µg/m³.

A comparison of the measured levels of pollutants in Tallinn to Swedish conditions show that in Tallinn the SO₂-levels are higher while the levels of NO₂ are comparable to those in Stockholm and Gothenburg. Regarding VOC, although there are not any exactly comparable measurements, some of the sites in Tallinn seem to be more exposed to VOC while others presumably would be comparable.

Although no regular annual measurements have been performed during these campaign measurements in Tallinn, the calculated average values from all campaigns would give a good indication of longtime average levels. In comparison to EU annual limit values it seems that for SO₂ the measured levels in Tallinn would be well below the limit value. For NO₂ the measured average concentrations at the most exposed sites were approximately 10% below the EU annual limit value for protection of health, but exceed an upper assessment threshold requiring measurements. There are presently no limit values within the EU regarding VOC. The expert group on air quality has however recently suggested an annual average for benzene of 5 µg/m³. This level would presumably be exceeded on an annual basis at several of the measurement sites in Tallinn.

From the results achieved it is obvious that the use of diffusive samplers, in combination with a careful selection of measurement sites, has been a fruitful strategy in mapping geographical variations in levels of air pollutants in Tallinn, an area of complex emission pattern. An important support in air quality management has been generated during a relatively short time period and at a reasonable cost. All results in combination have given an excellent insight into the variations of air pollution levels in the region as a whole as well as in different typical environments.