There are several advantages to developing a custom-made model for a project:
Can be constructed to address the project’s objectives and constraints
Directly tied to the CEM for the ecosystem
Excludes consideration of unrelated data and processes, which have no relevance to the metrics being forecast.
Avoids the cost and time required to collect unrelated data necessary to run some predictive models.
Makes use of existing data and local expertise of the ecosystem and be more likely to garner support from affected stakeholders and agencies if developed collaboratively.
Disadvantages to constructing a new model include:
The time and resources required for the development and testing of the model.
The availability of suitable expertise.
The effort required to verify, document, and certify and /or approve the model.
Danger of misspecification and bias, due to either omission of relevant explanatory variables or inclusion of falsely correlated variables in statistical models. (A safer approach is the use of models which simulate ecosystem processes, but these models require additional expertise for development.)
