Figure 2
From: Mechanisms for log normal concentration distributions in the environment
Numerical simulations (random sampling) of time-series concentration data and corresponding distributions, illustrating how different processes may influence concentration distributions. Panels (A)–(B): Log normally-distributed data (Eq. 3), e.g., influenced by exponential kinetics. Panels (C)–(D): Data reflecting the random fluctuation between three different log-normally distributed states (green, blue and red), yielding a multi-modal (mixture) distribution. Panels (E)–(F): Random mixing of the three distributions from panels (C)–(D), yielding a convolution distribution, tending towards a normal distribution. Panels (G)–(H): The log-normal distribution of panels (A)–(B), modulated by an oscillatory function (here sine function), illustrating the impact of, e.g. diurnal or seasonal cycles on observed distributions. The simulations were conducted using Matlab (ver. R2019b).
