AGC NZ 2023

GMDSI Workshop: Focus on the problem. Don’t let the model become the problem

There are many management contexts where numerical simulation can productively support environmental decision-making. This is not because it can replicate environmental behaviour very well. (We should drop that illusion.) However decision-focussed simulation can:

  • deliver information to decision-makers and stakeholders in ways that they can understand;
  • deliver the repercussions of lack of information to decision-makers and stakeholders in ways that they understand.

These are essential ingredients of decision-making.

Fundamental to the making of a decision is finding an answer to the question “what can go wrong if I do this or that”. Simulation can answer this question, while narrowing the range of predictive possibilities by harvesting all available information. It can be machine-learning on steroids – but only if it is designed to be so.

In this workshop we will talk about decision-support groundwater modelling in fresh and exciting ways.

During the morning, Jeremy White and John Doherty will discuss the latest technologies in model-partner software – software that can work with models to express what we know in ways that respect the limitations of our knowledge, and to harvest information from data while assessing risks that attend lack of information.

Then during the afternoon, we will talk about the modelling mindset that is needed to make this happen. Here we will question existing norms, and open up new possibilities. Most importantly, we will invite you to be part of the discussion. Catherine Moore and Eduardo de Sousa will present some common groundwater management problems. As a group, we will discuss how these problems can be decomposed into components that support information harvesting on the one hand, and risk assessment on the other hand.

Please join us – whether you are a modeller, or someone who is affected by decisions that are made using models.

There is no “right” way to model; that is for sure. However, more often than not, the slippery slide into unproductive modelling anguish begins when a model is commissioned before decomposing the problems that modelling must address, and identifying the information that modelling must harvest in order to address them. Our hope is that this workshop can help you put the horse back in front of the cart, by insisting that decision-support be the purpose of environmental modelling.


John Doherty
Jeremy White
Eduardo de Sousa
Catherine Moore

Monday 27 November 2023
8:30 AM – 5:00 PM NZDT
Catering supplied
Registrations closed