摘要:Gaseous elemental mercury was measured by two co-located continuous Tekran 2537B analyzers at the Atmospheric Station Kresin u Pacova, Czech Republic, from December 2012 to June 2013. The average concentration of 1.50 ng m-3 corresponds to the central European and northern hemisphere background levels of airborne mercury. The large dataset was further investigated by advanced statistical methods in order to characterize the variability of co-located measurement differences in 17 weekly data subsets. Internal variability in measurement data resulted from random effects only as proved by capability analysis. Due to considerable differences in statistical characteristics of individual weekly subsets a calculation of one conversion formula for the whole co-located measurement campaign in order to predict measure-ments by one instrument if measurements by the other are known was not a suitable option. Thus five clusters in which regression lines of weekly subsets showed similarities were identified. The cluster containing 4 weekly intervals from May-June 2013 performed the best with it's average regression equation y = 1.064x+ 0.006. Both analysers measured almost identically concen-trations up to 1.6 ng m-3 in this period.