Models may fit data but still fail to support reliable decisions. This audit checks whether assumptions, response memory, transformation uncertainty, and decision variables have been examined before the result enters a pumping, remediation, recovery, or subsurface-energy boundary.
A model can reproduce a groundwater response while leaving the mechanism, uncertainty pathway, or decision endpoint under-tested. The audit focuses on what changes when the same data are transformed into a decision.
The output is designed for internal technical use, client discussion, or the scoping stage of a larger project.
Review forcing history, response variables, simplifications, calibration targets, and validation gaps.
Identify which endpoint is exposed: allowable pumping, recovery time, remediation boundary, thermal margin, or uncertainty buffer.
Screen whether non-instantaneous response evidence is strong enough to justify memory-aware analysis.
Define how model-choice and transformation uncertainty should move into the decision variable.
Specify the smallest next analysis that can prove whether the decision changes materially.
Summarize findings in language usable by engineers, managers, and technical reviewers.
Clarify the decision, data type, exposure, and whether a paid diagnostic audit is justified.
Four to six weeks of data/model assumption review, decision-variable mapping, and memory relevance screening.
Compare conventional and memory-aware interpretation pathways on a selected site, dataset, or anonymized case.
Build a larger project around uncertainty propagation, reporting, review support, and possible publication.
Budget ranges should be matched to data access, confidentiality, reporting needs, and university contracting rules. A small diagnostic phase is usually a cleaner start than an oversized first project.