A Semiquantitative Groundwater Risk Screening Framework with Evidence Auditing and Conditional Scenario Perturbations in the Heihe River Basin

Groundwater screening under incomplete monitoring requires explicit separation of observations, qualitative judgments, and missing evidence. We examine a six-dimensional semiquantitative framework using three illustrative profiles labelled midstream Zhangye, Jiuquan–Jinta, and northern Qilian piedmont. An audit of zoning, monitoring, remote sensing, modelling, and management sources finds that none of the 18 original numerical assignments has a complete documented match among assessment boundary, reference period, observable indicator, and scoring rule. The empirical entries are therefore unresolved, while the original matrix is retained solely for methodological illustration. A missing-evidence interval rule prevents these gaps from being treated as observed middle categories. The illustrative baseline indices are 59.00, 62.00, and 43.25. Fixed score weight tests preserve their order, but independent groundwater status score changes can reverse it. Conditional scenario perturbations prescribe bounded score shifts rather than simulate hydrological responses. S1 lowers every index by 12.00 points by construction. S2 raises them by 9.75, 8.25, and 12.00 points; the second profile reaches 70.25 despite having the smallest increment. These results demonstrate how score assumptions, scale bounds, and class cutoffs affect interpretation. Empirical regional diagnosis requires matched observations, operational scoring rules, and independent validation.