A three day training course on model parameter estimation and uncertainty analysis using PEST.
We are pleased to announce that John Doherty of Watermark Numerical Computing, Australia, and author of PEST has agreed to come to the UK to present this one time event.
The training has been organised by Alastair Black of Groundwater Science Ltd UK following requests from the Environment Agency of England and Wales (EA) and a number of groundwater specialist consultancies.
The training will be joint led by John and Alastair.
We would like to thank ESI Ltd for volunteering to coordinate payments and booking.
What is PEST?
PEST (Parameter ESTimation) is an internationally recognized, state-of-the-art software package for model parameter estimation and model predictive uncertainty analysis. The training will mostly focus on the use of PEST in conjunction with MODFLOW. However, PEST is model-independent and can be used in conjunction with a wide variety of models – even simple EXCEL-based data-processing worksheets. In the UK, PEST is frequently used in model calibration, testing of conceptual models and quantifying model output sensitivity to various parameters. This practical course develops these methods. On the final day introduces additional techniques for quantification of predictive uncertainty in environmental models. It also describes how reduction of predictive uncertainty can be optimized through targeted acquisition of further data and how model-based decision-making can incorporate the information on predictive uncertainty that PEST provides.
The use of PEST in the UK has grown significantly over the last few years and this growth is expected to continue. This anticipation is a response to:
- current, increased use of models within the decision-making process;
- need for stakeholders to collectively explore the risks and benefits associated with different environmental / economic management options based on quantification of the uncertainty associated with those options; and,
- the availability of expanding computing power.
Who should attend the training?
Anyone with an interest in model calibration, and/or quantification of model parameter and predictive uncertainty. It would also benefit those who do not necessarily build models, but use models for decision-support, and would like to know what can be expected of models when used in that capacity.
Content:
Day 1 is comprised principally of lectures based on essential theory of parameter estimation and regularisation. Day 1 will also include UK based practical examples of previous PEST application. The benefits the techniques can bring to modelling assignments will also be discussed.
Day 2 is partly practical and partly theoretical. Delegates will have the opportunity to work through examples of MODFLOW model calibration using the regularisation techniques discussed the day before. There will also be time for clarification and further discussion on regularised inversion as applied to MODFLOW model calibration.
Day 3 focuses on uncertainty analysis, particularly as it relates to groundwater modelling. There will also be discussion on the best ways to include uncertainty quantification in model-based decision making and how uncertainty can be best reduced through targeted data acquisition.
Pertinent PEST features will be demonstrated. Workshops on the use of these utilities will be provided to participants to complete in their own time. Time will also be made available for more general discussions that participants may raise on perceived concerns/pitfalls of using automatic calibration tools in the environmental simulation processes. The need for decision-making to include an assessment of the uncertainties associated with different proposed management strategies can also be discussed.
All course participants will be provided with a memory stick which includes the latest version of PEST, literature on PEST, copies of all presentations and extensive workshop material for completion in their own time should they wish.
For additional information on PEST and its capabilities please see http://www.pesthomepage.org/Home.php
Further information and booking
For queries regarding the training, content, partial attendance or accommodation/travel please contact a.black@gwscience.co.uk Tel: +44 (0) 7951 495 121
For additional information on PEST or alternative training courses outside of the UK John can be contacted at johndoherty@ozemail.com.au, Watermark Numerical Computing, 336 Cliveden Avenue, Corinda 4075, Australia. Tel: 07 3379 1664
For booking please contact ESI: New Zealand House, 160 Abbey Foregate, Shrewsbury. SY2 6FD. Tel: +44 (0) 1743 276100, Fax: +44 (0) 1743 248600, Email: coursesuk-esi@esinternational.com.
The cost of the 3 day event is £750+VAT
NOTE: Delegates should bring their own laptops for hands on sessions.
ESI can accept payment by credit card over the phone from most major credit and debit cards including MasterCard, Visa, Maestro, JCB, Visa Electron and Solo (we do not accept American Express or Diners ClubCard).
We look forward to seeing you!