Groundwater Modelling Uncertainty Workshop

Canberra - 22nd July 2019

A 1-day Groundwater Modelling Uncertainty Workshop, hosted by the National Centre for Groundwater Research and Training (NCGRT) and sponsored by BHP and Rio Tinto was held in Canberra on Monday 22nd July 2019. This highly successful event was attended by over 100 participants from government, academia and private industry. It provided an engaging forum to explore the issue of model predictive uncertainty. 

Analysis of predictive uncertainty is central to the assessment of risk. Hence it is integral to environmental management and decision-making. The workshop explored software and methodologies that can evaluate the uncertainties associated with predictions made by environmental models, both before and after they have been calibrated. It also addressed the important issue of how complex a groundwater model actually needs to be in order to provide adequate decision support, and how its specifications should reflect availability (or lack of availability) of data.

Presenters from government, academia and industry discussed decision-support modelling from their own perspectives. Included in their discussions were their perceptions of its current strengths and weaknesses.  The workshop was addressed by international experts on model uncertainty quantification, including John Doherty (author of PEST), Jeremy White (author of PEST++) and Ty Ferre (University of Arizona and 2017 Darcy Lecturer). Presenters from industry shared their valuable experiences in applying state-of-the-art methods of modelling and model uncertainty quantification in a wide variety of real-world settings.

Overall, the mix of approaches, experiences and personalities created a unique learning and sharing environment from which all attendees could benefit. The NCGRT would like to thank all participants for their interest, and for their meaningful contributions to the event.

A compilation of presentations from the workshop can be found here.