Inter-city and inter-subject variability in the airborne pollen concentrations or counts due to random interclass effects.

a Number of monitoring stations or human subjects.b Intercept variances were calculated by the mixed models assuming the monitoring stations or the human subjects as random effects and the measurement days as fixed effects. The natural-log-transformed pollen concentrations (pollens m-3) or counts (pollens cm-2 day-1) were used for the mixed models.c Geometric coefficient of variation (CV) was calculated according to eq (1).Inter-city and inter-subject variability in the airborne pollen concentrations or counts due to random interclass effects.