Modeling, Environmental

Abstract

Abstract In discussing environmental models, it is useful to distinguish between process‐based or mechanistic models and phenomenological or empirical models. Although the distinction is not always clear, mechanistic models are based on a scientific theory about the underlying environmental processes, while empirical models are based on observation alone. This article focuses on statistical analysis under mechanistic environmental models. This is not to imply that empirical environmental models are less interesting or less useful. Rather, it is because empirical environmental models are, from a statistical perspective, no different from empirical models in other fields. While the methods of statistical analysis under mechanistic environmental models are certainly not unique, such models tend to share certain characteristics that determine in part the kind of statistical issues that arise, and even the kind of statistical methods used to address these issues.