When can a Computer Simulation act as Substitute for an Experiment? A Case-Study from Chemisty

Publication Year:
Usage 167
Downloads 167
Social Media 1
Tweets 1
Repository URL:
Arnold, Eckhart; Kästner, Johannes
Most Recent Tweet View All Tweets
preprint description
In this article we investigate with a case study from chemistry under what conditions a simulation can serve as a surrogate for an experiment. We set out with a brief discussion of the similarities and differences between simulations and experiments. There are three fundamental differences: 1) Ability (of experiments) to gather new empirical data. 2) Ability to operate directly on the target system. 3) Ability to empirically test fundamental hypotheses. Given that there are such fundamental differences it becomes an important question if and under what conditions simulations can still act as surrogate for experiments. We investigate this question by analysing a simulation of H2 -formation in outer space. We find that in this case the simulation can act as a surrogate for an experiment, because there exists comprehensive theoretical background knowledge about the range of phenomena to which the investigated process belongs and because any particular modelling assumptions as, for example, on the validity of approximations, can be justified. If these requirements are met then direct empirical validation of a “virtual experiment” may even be dispensable. We conjecture that in the absence of comprehensive theoretical background knowledge direct empirical validation of “virtual experiments” remains unavoidable.