MODEL CALIBRATION Inversion Auto Calibration with UCODE

May 16-18, 2008

which is intended to introduce ground-water professionals to inverse modeling concepts and their use via UCODE, relying heavily on hands-on exercises for automatic calibration of ground-water models to promote understanding of UCODE and avoid "black-boxing."

Course sessions are from 8:00 AM to 5:00 PM.

We begin with a brief review of calibration concepts, and introduce inversion theory, then follow with description of the input, output and special topics associated with using UCODE. We use hands-on exercises to demonstrate the application of UCODE. Each exercise involves more complex issues such that the attendees will be exposed to use of all of the available features of UCODE:

• checking problem set-up;
• developing an objective function surface;
• estimating parameters;
• evaluating statistics for:
  a) identifying inadequate data and parameters that probably cannot be estimated,
  b) evaluating estimated parameter values,
  c) evaluating the accuracy of the model's representation of the system, and
  d) quantifying the uncertainty associated with predictions made using the calibrated model
• procedures for inputting complex parameterizations and simulated-equivalents; as well as
• using the peripheral output files of UCODE, to
  • determine the validity of your model,
  • guide further work related to data collection and simulation, and
  • present the certainty of your results.
  
  Potential pitfalls are discussed

UCODE_2005

UCODE_2005 is a universal inversion code, that performs inverse modeling, posed as a parameter-estimation problem, using nonlinear regression. UCODE is not limited to inverting ground-water problems, but the course is presented as an approach to ground-water model calibration. UCODE performs inverse modeling, and was developed to:

• manipulate application model input files and read values from application model output files;
• compare user-provided observations with equivalent simulated values derived from the values read from the application model output files using a weighted least-squares objective function;
• use a modified Gauss-Newton method to adjust the value of user selected input parameters in an iterative procedure to minimize the value of the weighted least-squares objective function;
• report the estimated parameter values; and
• calculate and print statistics that reflect the quality of the model and its predictions.

Any process model or set of models can be used; the only requirements are that models have numerical (ASCII or text only) input and output files, that the numbers in these files have sufficient significant digits, that all required models can be run from a single batch file or script, and that simulated values are continuous functions of the parameter values. Process models can include pre-processors and post-processors as well as one or more models related to the processes of interest (physical, chemical, and so on), making UCODE_2005 a powerful calibration and uncertainty analysis tool. An estimated parameter can be a quantity that appears in the input files of the process model(s), or a quantity used in an equation that produces a value that appears in the input files. In the latter situation, the equation is user-defined. UCODE_2005 can compare observations and simulated equivalents. The simulated equivalents can be any simulated value written in the process-model output files or can be calculated from simulated values with user-defined equations. The quantities can be model results, or dependent variables. For example, for ground-water models they can be heads, flows, concentrations, and so on. Prior, or direct, information on estimated parameters also can be considered. Statistics are calculated to quantify the comparison of observations and simulated equivalents, including a weighted least- squares objective function. In addition, data-exchange files are produced that facilitate graphical analysis.

UCODE_2005 can be used fruitfully in model calibration through its sensitivity analysis capabilities and its ability to estimate parameter values that result in the best possible fit to the observations. Parameters are estimated using nonlinear regression: a weighted least-squares objective function is minimized with respect to the parameter values using a modified Gauss-Newton method or a double-dogleg technique. Sensitivities needed for the method can be read from files produced by process models that can calculate sensitivities, such as MODFLOW-2000, or can be calculated by UCODE_2005 using a more general, but less accurate, forward- or central-difference perturbation technique. Problems resulting from inaccurate sensitivities and solutions related to the perturbation techniques are discussed in the course. Statistics are calculated and printed for use in (1) diagnosing inadequate data and identifying parameters that probably cannot be estimated; (2) evaluating estimated parameter values; and (3) evaluating how well the model represents the simulated processes.

Results from UCODE_2005 and associated codes RESIDUAL_ANALYSIS and RESIDUAL_ANALYSIS_ADV can be used to evaluate how accurately the model represents the processes it simulates. Results from LINEAR_UNCERTAINTY can be used to quantify the uncertainty of model simulated values if the model is sufficiently linear. Results from MODEL_LINEARITY and MODEL_LINEARITY_ADV can be used to evaluate model linearity and, thereby, the accuracy of the LINEAR_UNCERTAINTY results.

UCODE_2005 can also be used to calculate NONLINEAR confidence and predictions INTERVALS, which quantify the uncertainty of model simulated values when the model is not linear. CORFAC_PLUS can be used to produce factors that allow intervals to account for model intrinsic nonlinearity and small-scale variations in system characteristics that are not explicitly accounted for in the model or the observation weighting.

UCODE_2005 and the six associated codes are intended for use on any computer operating system. The programs consist of algorithms programmed in Fortran 90/95, which efficiently perform numerical calculations. The model runs required to obtain perturbation sensitivities can be performed using multiple processors.

This course is designed for the hydrogeologist and engineer familiar with ground-water flow modeling who would like to learn to use UCODE to calibrate their models. Participants will benefit most from this short course if they have a working knowledge of ground-water modeling and some knowledge of basic statistics.

Attendees will receive course notes; the UCODE software and associated USGS reports, including the UCODE manual; and data files for example problems.

Dr. Eileen Poeter is Director of the IGWMC and a Professor at Colorado School of Mines. She has been modeling in the academic and geohydrologic consulting arena for 32 years and teaching modeling as semester and short courses for 25 years. Dr. Poeter is the first author of UCODE. She has authored over 100 articles and reports in the area of ground water with specialization in ground-water modeling, evaluation, and simulation of heterogeneity, use of soft data such as geophysical data, and using inverse modeling as a tool to improve such endeavors.

ENROLLMENT & REGISTRATION

The fee for the short course is discounted for those who are registered for the 2008 MODFLOW 2008 to $1045 before May 2 ($1245 after May 2). The regular course fee for those who are not attending the conference is $1195 before May 2 ($1395 after May 2) . You will be invoiced for the tuition balance. The fee covers instruction, course notes, refreshments, and continuing education units. Lodging and meals are not covered. Payment in full must be received prior to the beginning of the course. The registration fee may be paid by check, VISA, MasterCard, AMEX or purchase order. A maximum of 20 applicants will be able to attend the course due to space limitations.

If cancellation is made before May 2, 2008, tuition will be refunded except for the $100 deposit. Cancellations made after that date will forfeit a $150 processing fee in addition to the $100 deposit. In case of cancellation, the right to course material is forfeited. Substitutions are permitted at any time prior to the start of the course. Please be advised that if the program is cancelled due to insufficient enrollment, Colorado School of Mines will return all course fees, but will not be liable for any other expenses incurred by the prospective attendees.

To register click here. We will then contact you with more information based on your payment selection.

COURSE LOCATION

The course will be held in the Berthoud Hall, Rm. 201, at the Colorado School of Mines in Golden, Colorado. A city and campus map will be sent to you upon course enrollment.

TRANSPORTATION AND LODGING

The Colorado School of Mines is located in Golden, which is 20 minutes west of Denver, 1 hour west of the Denver International Airport and can be reached by taxi, airport shuttle or rental car. Lodging is available in Golden, within walking distance of the campus. Information regarding transportation, parking, and lodging will be sent upon registration.

For more information, contact:
International Groundwater Modeling Center
phone: +1 303 273-3103
fax: +1 303 384-2037
email: igwmc@mines.edu

IGWMC MAIN PAGE

Last Update: January 9, 2008