Bounce Around Ideas for Free
The GMDSI philosophy of decision support modelling is expressed in its decision support modelling manifesto. Briefly:
- Design and build a model with a prediction in mind
- Quantify the uncertainty of that prediction
- Reduce that uncertainty through model-based data assimilation.
This requires compromise.
A model cannot be too complex, because complex models take too long to run and can throw numerical tantrums. On the other hand, simple models may be challenged when it comes to quantifying and reducing the uncertainties of some decision-critical model predictions.
Compromise comes at a cost. The cost must be included in predictive uncertainty intervals.
Once an appropriate modelling strategy has been figured out, the hard work has only just begun. How should the model be parameterized? How should history-matching and uncertainty analysis be accomplished? What software should be used? How should it be linked to the model?
We can’t do your job for you, but we are happy to offer some free advice – over the internet. Here is what we would like to do. Let us know if you are interested.
- We can meet for two hours – two GMDSI personnel with you and your modelling team. Tell us your story.
- We’ll think about it for a while – maybe even ask you a few more questions by email.
- In another week we’ll meet again. Then we can shoot the breeze together on possible modelling approaches, and perhaps some implementation details.
- We’ll write a short report (two to three pages) as a record of our conversations and thought processes, and to make sure that what we say is clear.
Contact us if you are interested, and we will take it from there.
Banner image: BHP