A non-instantaneous response is easy to dismiss as a detail of model fitting. It becomes harder to dismiss when the decision itself is temporal.

Examples:

  • when a dewatering plan must know how quickly drawdown reaches a control boundary;
  • when a groundwater reserve plan must know how long recovery takes after emergency pumping;
  • when a TRT interpretation must decide which part of the thermal response represents the formation rather than the borehole.

Groundwater memory is not about adding a lag parameter for its own sake. It asks whether flux, head, boundary, deformation, or thermal response moves out of step in a way that changes the action.