Modelling guidelines––terminology and guiding principles

摘要: Abstract Some scientists argue, with reference to Popper’s scientific philosophical school, that models cannot be verified or validated. Other and many practitioners nevertheless use these terms, but very different meanings. As a result of an increasing number examples model malpractice mistrust the credibility models, several modelling guidelines are being elaborated in recent years aim improving quality studies. This gap between views lack consensus experienced community strongly perceived need for commonly agreed is constraining optimal benefits models. paper proposes framework assurance guidelines, including consistent terminology foundation methodology bridging philosophy pragmatic modelling. A distinction made conceptual model, code site-specific model. subject confirmation falsification like theories. may within given ranges applicability accuracy, it can never universally verified. Similarly, validated, only applications pre-specified performance (accuracy) criteria. Thus, model’s validity will always limited terms space, time, boundary conditions types application. implies continuous interaction manager modeller order establish suitable accuracy criteria predictions associated uncertainty analysis.