SILVERPIT DEBATE - MANAGING UNCERTAINTY IN COMPLEX MODELS
Quentin Cooper explores scientific developments, with mathematics under the spotlight.
SILVERPIT DEBATE
About 65 million years ago, an object 120 metres wide and weighing 7 million tonnes fell from space. It plunged into the shallow sea off the coast of what is now Yorkshire. It left a 3km crater, and showered havoc and destruction on the surrounding area.
Or so many geologists think.
In 2002, Phil Allen published the discovery of what was hailed in many quarters as the UK’s first and so far only impact crater. Since then, some have cast doubt over its origins.
Could the strange, ring-shaped fault lines surrounding the Silverpit Crater, as it came to be known, point to a more earthly genesis?
The debate has rumbled on ever since, and recently the Geologist leading the case for the opposition has presented what he believes to be further evidence supporting his subterranean theory.
Professor John Underhill of Edinburgh University is joined by the Silverpit’s discoverer, Phil Allen of Production Geoscience Ltd to thrash it out with Quentin.
MANAGING UNCERTAINTY IN COMPLEX MODELS
In many branches of science, more and more predictions are made by running computer models.
But as these models get increasingly complex, a better understanding of the mathematics of uncertainty is becoming essential, especially when important policy decisions hinge on the scientists’ predictions.
From flu outbreaks to climate change, from bridges and plumbing, scientists have to assess the degree of uncertainty by comparing different predictive models.
The MUCM (Managing Uncertainty in Complex Models) project has been set up to draw together a set of statistical rules and technologies that could impact on all areas of science for the better.
Leader of the £2 million project, Professor Tony O’Hagan of Sheffield University’s Department of Probability and Statistics and Dr Peter Challenor, Head of Ocean Observations and Climate Research Group at the University of Southampton discuss the odds of success.