Accuracy limitations and the measurement of errors in the stochastic simulation of chemically reacting systems (Q2576282)

The authors consider a chemically reacting system involving a finite number of species and of reactions. The dynamics of this system is supposed to be simulated using a stochastic formulation. The authors consider the Schlögel reaction as a model. The corresponding chemical master equations are solved using the \(\tau\)-leaping method. In order to evaluate the efficiency of Monte Carlo simulations, when comparing to experiments, the authors define two distances: the Kolmogorov one and the histogram one. Using these two distances, the authors build some so-called self distances between two sets of Monte Carlo samples which satisfy some properties. The main part of the paper gives upper bounds of these self distances mainly in terms of the sizes of the samples. The paper ends with some numerical experiments which validate these upper bounds.