Types of Models

Statistical Models – are particularly useful when data are available; however, because they do not incorporate explicit causal mechanisms, they are vulnerable to error when attempts are made to extrapolate forecasts to alternatives that fall outside the range of the analyzed data.

Mechanistic Process Models – use explicit numerical descriptions of causal mechanisms within the ecosystem, and so are more grounded in scientific knowledge of ecosystem dynamics. Process models can be expensive to develop, difficult to calibrate, and may require expertise beyond the reach of some planning units.

Hybrid Models – are semi-empirical models that combine statistical and mechanistic elements within a single algorithm and have increased in popularity because they make use of the full range of available knowledge and modeling techniques within a single application.