This modelling component of the SYMPLE web site has been prepared with the help of John Doherty. John is the author of PEST and associated software. Until recently, he was also heavily involved in production of training material and videos for GMDSI.
John also helps with SYMPLE training. He has conducted courses (personal and web-based) on parameter estimation and uncertainty analysis, as well as a course on MODFLOW 6 input and output protocols. He is also a member of the SYMPLE scientific committee.
John has had many years of consulting experience (and a few in academia). As those who know him are aware, he holds rather strong views on decision-support groundwater modelling. Decision-support modelling is NOT about creating a digital replica of a complex, unseen and highly heterogeneous natural system. This is not possible.
Decision-support modelling is all about harvesting and delivery of information to decision-makers and decision-stakeholders. This information resides in data that emerges from site characterization, and in historical and recent measurements of a system’s behaviour. Harvesting and delivery of information doesn’t happen automatically. It must be planned. This is called “problem decomposition” – an essential prerequisite to decision-support modelling.
Just as importantly, decision-makers and decision-stakeholders need to be aware of the consequences of information insufficiency. This is expressed as predictive uncertainty. With uncertainties elaborated, decision-making can become risk-aware.
