Tie the model to a decision
Define the question, spatial boundary, time horizon and response variable. Ask what difference in the output would change the project decision. This prevents unnecessary model complexity and exposes situations where the required conclusion is more specific than the available evidence.
Separate calibration from evaluation
Record which observations informed parameter selection and which provide an independent check. Review how errors vary across conditions rather than relying on a single summary score. A fit to familiar conditions does not automatically establish reliability under a new operating or climate regime.
Distinguish scenarios and forecasts
Describe the assumptions behind each scenario, including boundary conditions and non-climate drivers. Use sensitivity analysis to identify influential inputs. Where probabilities cannot be supported, show conditional outcomes or ranges instead of presenting the central curve as a prediction.
Report the usable conclusion
State what the model supports, where extrapolation starts and what would alter the conclusion. Deliver the assumptions register, comparison plots and decision implications together. The reader should be able to identify both a robust finding and a fragile ranking.
Sources & context
Background reading checked 9 October 2026. The review questions and structures above are BGF editorial guidance; the references are not BGF project evidence or endorsements. Apply them to the relevant site, data and agreed scope.