RANDOMNESS IN AIRBORNE POLLEN - CHAOS OR NOISE

The most abundant biological particles in the atmosphere are pollen gr ains and spores. Previous analyses of the dynamic characteristics of a tmospheric pollen time series based on the study of the correlation di mension indicate that the system can be described by a low dimensional chaotic map. As the determination of a fractal dimension is not compl etely reliable, a less ambitious but secure task is to find the nonlin ear character of the time series. A recently developed method to disti nguish nonlinear deterministic systems from linear stochastic systems is applied to a pollen time series of about 8000 values. The nonlinear structure of the series is obtained.