A new multistage groundwater transport inverse method: Presentation, evaluation, and implications

More computationally efficient methods of using concentration data are needed to estimate groundwater flow and transport parameters. This work introduces and evaluates a three-stage nonlinear-regression-based iterative procedure in which trial advective-front locations link decoupled flow and transport models. Method accuracy and efficiency are evaluated by comparing results to those obtained when flow- and transport-model parameters are estimated simultaneously. The new method is evaluated as conclusively as possible by using a simple test case that includes distinct flow and transport parameters, but does not include any approximations that are problem dependent. The test case is analytical; the only flow parameter is a constant velocity, and the transport parameters are longitudinal and transverse dispersivity. Any difficulties detected using the new method in this ideal situation are likely to be exacerbated in practical problems. Monte-Carlo analysis of observation error ensures that no specific error realization obscures the results. Results indicate that, while this, and probably other, multistage methods do not always produce optimal parameter estimates, the computational advantage may make them useful in some circumstances, perhaps as a precursor to using a simultaneous method.