SHORTCOURSE: Model Sensitivity Analysis, Calibration and Uncertainty Evaluation

When:
January 13, 2020 – January 14, 2020 all-day
2020-01-13T00:00:00-08:00
2020-01-15T00:00:00-08:00
Where:
UC Davis-Walter A. Buehler Alumni Center
530 Alumni Lane Davis
CA 95616

The Sustainable Groundwater Management Act calls out explicitly for model evaluation and for the evaluation of uncertainty in the models that we develop for the Groundwater Sustainability Plans. Directly from the BMP on modeling, the Department of Water Resources suggests:

“Sustainable groundwater management and policy decisions must be based on knowledge of the past and present behavior of the surface and groundwater system, the likely response to future changes, and the understanding of the uncertainty in those responses.”

The focus of this course is on methods for sensitivity  analysis, data assessment, model calibration, and evaluation of model uncertainty.  The course will introduce these topics and use basic results to provide guidelines for how to use these methods in practice. The course will emphasize  spatially-discretized systems and “snap-on” tools that are not model or code specific (e.g. UCODE_2014 and PEST). Insights and examples of both local and global methods will be presented through hands-on use of local methods in UCODE_2014 and PEST, and global methods  (Morris and Sobol). Basic concepts covered include:

  • Analyze data to be used for model development. Based on knowledge of the system, decide how to define parameters.
  • Determine parameter values that provide a best fit to observations.
  • Calculate predictions and obtain measures of prediction uncertainty.
  • Using the initially constructed model, identify parameters important to observations, parameters important to predictions, and observations important to predictions. We will
    •  Question whether these simulated relationships are consistent with real-world conditions.
    • Assess how much information the data provide and what level of model complexity can be supported by the data
    • What potential new data should be collected to improve predictions

Exercises will be provided and used throughout the course and large-scale examples will be highlighted to show how these methods are used in practice. These techniques will be discussed using groundwater model examples but can be applied to other models, and example of applications in different fields will be provided.

Click here for more information and to register.