|
|
|
          
|
ADVective-Transport Observation (ADV) Package,A Computer Program for Adding Advective-Transport Observations of Steady-State Flow fields to the Three-Dimensional Ground-Water Flow Parameter-Estimation Model MODFLOWP
ABSTRACT Observations of the advective component of contaminant transport in steady-state flow fields can provide important information for the calibration of ground-water flow models. This report documents the ADV, Advective-Transport Observation, Package, which allows advective-transport observations to be used in conjunction with the hydraulic head and head-dependent flow observations included in the three-dimensional ground-water flow parameter-estimation model MODFLOWP. The particle-tracking routine used in the ADV Package duplicates the semi-analytical method of MODPATH, as shown in a sample problem. Users are allowed to track particles in a forward or backward direction, or even to include effects such as retardation through manipulation of the effective-porosity value used to calculate velocity. Although effective porosity could be included as a parameter in the regression, this capability is not included in this package. The weighted sum-of-squares objective function, which is minimized in the parameter estimation process, was augmented to include the square of the weighted x-, y-, and z-components of the differences between the simulated and observed advective-front locations at defined times, thereby including the direction of travel as well as the overall travel distance in the calibration process. The sensitivities of the particle movement to the parameters needed to minimize the objective function are calculated for any particle location using the exact sensitivity-equation approach; the equations are derived by taking the partial derivatives of the semi-analytical particle-tracking equation with respect to the parameters. The ADV Package is verified by showing that parameter estimation using advective-transport observations produces the true parameter values in a small but complicated test case when exact observations are used. To demonstrate how the ADV Package can be used in practice, a field application is presented. In this application, the ADV Package is used to investigate the importance of advective-transport observations relative to head-dependent flow observations when either or both are used in conjunction with hydraulic-head observations in a simulation of the sewage-discharge plume at Cape Cod, Massachusetts. The analysis procedure presented can be used in many circumstances to evaluate the probable effect of new observations on the parameter estimation. The procedure consists of two steps: (1) Parameter sensitivities and correlations calculated at initial parameter values are used to assess the model parameterization and expected relative contributions of different types of observations to the regression; and (2) optimal parameter values are estimated by nonlinear regression and evaluated. In this application, advective-transport observations improved the calibration of the model and the estimation of ground-water flow parameters, and use of formal parameter-estimation methods and related techniques produced significant insight into the physical system.
The following figure shows how a simulated particle path is compared to its expected location using discrepancies in three coordinate directions. Sensitivities are calculated for the movement of the particle in these three coordinate directions, and can be used to determine which parameters are most important to advective transport. This is accomplished using the prediction scaled sensitivites described in the report 'Methods and Guidelines for Effective Model Calibration'.
ADV is distributed with MODFLOWP. The program and documentation can be downloaded from the USGS Ground Water software repository.
mchill@usgs.gov Last Modified: March 17, 1999 |
