Upcoming Events: Classes Mary C. Hill Ground-water model calibration and uncertainty analysis (A Course for Geologists and Engineers) Fall, 2000 University of Colorado Civil, Environmental and Architectural Engineering Boulder, Colorado Course Description Models are used extensively to evaluate ground-water systems and to predict their response to such things as changes in pumpage and proposed remediation efforts. Because many aspects of ground-water systems are unknown, most models are calibrated. Calibration commonly is achieved by trial and error alone, but these methods provide less insight than is possible. This course teaches how sensitivity analysis, nonlinear regression, and associated statistics can be used to greatly improve how data is used to calibrate and test ground-water models. For example, parameters that can not be estimated accurately and uniquely with the available data can be quickly identified. Parameter values that produce the best fit between simulated and observed hydraulic heads, concentrations, and so on can be estimated by nonlinear regression. Measures of prediction uncertainty and measures of the importance of existing and potential observations are a natural consequence of regression methods. Prerequisites: Basic statistics, computer usage. Ideas are taught using ground-water modeling but apply to any modeling; ground-water model experience is not a prerequisite. Teaching staff Mary C. Hill Research scientist, U.S. Geological Survey, Boulder, CO. Author of MODFLOWP, the popular PCG2 solver for MODFLOW, MODFLOWP, and MODFLOW-2000, articles on the numerical methods of solvers, nonlinear regression, confidence intervals, and calibration methodology, and co-author of UCODE and MODFLOW-2000. Dr. Hill has conducted and consulted on numerous national and international ground-water investigations, including saltwater intrusion, ground-water supply and stream interaction, and evaluation of the Death Valley regional ground-water flow system, which underlies the U.S. proposed high-level nuclear waste site at Yucca Mountain. Dr. Hill has taught for 19 years and holds a Ph.D. in Civil Engineering, Princeton University. Richard L. Cooley Research hydrologist, U.S. Geological Survey, Lakewood, CO. Authored some of the first publications on applying nonlinear regression to calibration and uncertainty analysis of ground-water models. Awarded the O.E Meinzer award of the Geological Society of America for his 1979 paper on the subject. Primary fields of research have been numerical methods, calibration, and uncertainty analysis of ground-water models; he has published on a variety of topics in hydrogeology, including variably saturated flow, geomorphology, and recharge through desert alluvial fans. Dr. Cooley has taught for 30 years and holds a Ph.D. in Geology, Pennsylvania State University. Claudia C. Faunt Hydrologist at the U.S. Geological Survey in San Diego, California. Expert in the development of hydrogeologic models for ground-water model development using advanced three-dimensional data base and visualization methods. Dr. Faunt has used these methods extensively to analyze the Death Valley regional ground-water flow system in Nevada and California. Dr. Faunt holds a Ph.D. in Geological Engineering, Colorado School of Mines. Program Schedule Course participants will learn nonlinear regression methods using the U.S. Geological Survey computer programs MODFLOW-2000, an inverse ground-water flow model that is numerically sophisticated but applies to limited situations, UCODE, a universal inverse code that can be used with any model, and MT3DMS, a forward ground-water transport program. These public domain programs are well-documented, tested, and suitable for complex field application. Class hours will be from 12:30 to 1:45, Tuesday and Thursday, starting August 29. The course will proceed as follows: Ground-water management problem used for class exercises Review of the ground-water flow and advective-transport capabilities of MODFLOW-2000 Presentation of graphical interfaces for MODFLOW-2000 Methods for model calibration Model calibration using nonlinear regression Lectures and exercises using MODFLOW2000 Defining parameters and steady-state observations Calculating sensitivity-equation sensitivities Evaluating data and parameters using dimensionless and composite scaled sensitivities Performing parameter estimation Diagnosing parameter-estimation convergence problems Multiple linear regression Multiple nonlinear regression Using a two parameter version of the test problem to investigate how the modified Gauss-Newton method works Nonlinear regression Parameter uncertainty and correlation Evaluation of model fit Prediction uncertainty Testing for model nonlinearity Predictions and prediction uncertainty Effect of potential new transient observations Model calibration using transient observations Lectures and exercises using UCODE Calculating sensitivities by perturbation Investigating the contribution of concentration observations using UCODE and MT3DMS, and presentation of associated graphical interfaces Used hydrogologic data to constrain ground-water model development, methods and examples - Dr. Claudia C. Faunt Field example: Mirror Lake, New Hampshire Field example: Modeling the Death Valley regional ground-water system Composite scaled sensitivities indicate parameters for which the data provide substantial information (large values) and little information (small values) Optimal remediation strategies under parameter uncertainty Model nonlinearity and sensitivity analysis, regression, and uncertainty analysis - Dr. Richard L. Cooley Parallel computing Parameter definition using interpolation methods such as kriging mchill@usgs.gov Last Modified: August 15, 2000