Small Scale Temporal-Spatial Variability of Particle Number Concentration in Augsburg, Germany

Abstract

P-011 Introduction: Epidemiological time-series studies on air pollutants and human health typically use either air pollutants concentrations data obtained at one central monitoring site or averaged concentrations data obtained at several monitors across the city. If the measurements at monitors at different locations are not highly correlated, there is a significant potential for exposure misclassification in these studies. Until now only one study reports continuously measurements of particle number concentrations (NC) across an urban area for 14 days (Buzorius et al., 1999). A new fixed monitoring site to provide hourly particle measurements for future epidemiological studies was planned 2004 in a medium-sized European city. Before establishing the new monitoring site we investigated the representativeness of the selected measurement site for the urban area. Methods: The study was conducted in December 2003 (winter period) and April-May 2004 (spring period). In the winter period NC were measured at three urban background sites (two sites were moderately influenced by traffic) and one remote site for 12 days, whereas the measurements in the spring period were limited to the three urban sites only (38 days). The measurements were conducted by means of Condensation Particle Counters 3025 and 3022. Results: In general, the NC time series measured in different places show high correlations both on hourly as well as on daily base. The Spearman correlation coefficients between the NC in the winter period range from 0.77 to 0.92 on hourly base and 0.85 to 0.98 on daily base. The associations between the different urban locations are somewhat weaker in the spring period, but still strong (0.81 – 0.83 on hourly base and 0.85 – 0.86 on daily base). The associations are influenced by wind direction, whereas the influence of other meteorological parameters (wind speed or temperature) is less prominent. The highest NC in the spring period were measured at the two background site with traffic impact (18600 cm-3 or 16904 cm-3) followed by the mere background site (10111 cm-3). Discussion and Conclusions: The selected measurements site is representative with respect to the monitoring of the urban aerosol in the investigated city area. The correlations between the NC measured at different locations in an urban area are relatively high. It indicates that the particle number concentrations are related to traffic intensity, which is the main aerosol source relatively homogenous distributed across the city. On Sundays the correlations are lower than on working days, when traffic intensity is relatively high.