Workshops

The second day of the conference will be split into seven parallel sessions. Each session will consist of three talks on the area of interest followed by a chaired discussion.

The aim of these sessions is to allow people to attend the session which is of most relevance/interest to them and therefore gain the most from the second day. Each session will aim to move a specific topic area forward and highlight active research areas. Abstracts are welcomed to each area as the aim of these sessions is to gain as broad an input as possible. When submitting your abstract please indicate via the check boxes which session you would like your abstract considered for.

The seven sessions are outlined below

Communicating Spatial Uncertainty

There are numerous ways to communicate spatial uncertainty (shading, contours, bubbles etc.), however, there is no recognised best practice. The aim of this session is to explore the problems associated with particular methods, such as the misunderstanding of hard lines, and to examine methods for determining a common approach.

Decision Making Under Uncertainty

Decision makers are constantly faced with the requirement to make decisions with incomplete supporting information. This can be because it is impossible to obtain all of the information known to be relevant (known unknowns). It can also be because it is not known what information is relevant (unknown knowns and unknown unknowns). Recent work has sought to address these issues via methods such as Hierarchical Process Modelling based upon 'Italian flag' assessments. The aim of this session is to review and discuss the application of this and other methods to support decision makers.

Representing Uncertainty and Risk in Cost Estimates for Complex Systems

Cost estimates for large and complex projects and financial estimation for large systems pose significant challenges. These challenges may come in the form of managing engineering and innovation challenges, business change, or political stakeholders. This workshop will explore how different perspectives on uncertainty may influence approaches to its management in these different contexts. As a by-product of facilitating the sharing of different approaches to these challenges attendees will explore the differences between uncertainty and risk, and can expect to take away some new thinking used to translate uncertainty into cost estimates and financial projections.

Uncertainty Propagation and Extrapolated Model Predictions

Physically, phenomenologically and empirically derived models of complex systems provide the potential for efficient prediction and control. The challenges of uncertainty quantification and calibration of individual models within the domain of the data have attracted considerable effort and research, and well-developed solutions are available. However, increasingly models are being used (i) in combination, with the resultant need to propagate uncertainty between models, and (ii) to predict and control systems beyond the tested input conditions. The aim of this session is to explore methodology: for accurate calculation and representation of uncertainties across modelling chains; for representation and communication of uncertainties in extrapolated predictions; and for combining accelerated and conventional data.

Public Perception of Risk

There is a need across government and industry to enable the public to understand the risks posed by various processes, activites and phenomena. This enables the public to decide whether to take medication (E.g. statins and vaccinations) and ensure they are informed of potential risks posed. There is also a need to understand how the public will react to, particularly unfamiliar risks (e.g. space weather), to ensure that processes are put in place to support this. The aim of this workshop is to explore methodologies for communicating uncertainty generally and methodologies for understanding potential reactions.

Uncertainty Modelling for Large Scale Computer Models

Large scale computer models pose considerable challenges to the calculation, propagation and extrapolation of uncertainty. However, in these computationally expensive problems, the complexity of the system almost uniquely precludes the full calculation of this uncertainty – we can never hope to resolve the models across the entire modelling domain - while making the presentation of uncertainty associated with the prediction crucial. This workshop will explore the use of ensemble modelling, calibration and forecasting through applications in climate and meteorological science. It will focus specifically on problems where model uncertainty cannot be tractably calculated and therefore alternative methodology is required. Methodology will be explored for complex uncertainty propagation through multi-model systems and will also look at how this uncertainty can be used to provide risk based estimates of potential hazards.

Planning for Complex Futures

Decision makers are required to make complex investment decisions, the impact of which will take a long time to realise, with limited ability to adjust when the future reality does not match the assumed future that the decisions were based on. One basis for supporting these decisions is via the use of scenarios, where concrete predictions of possible futures are defined and plans are produced to meet those possible futures. However, the one thing that is guaranteed about the future is that predictions about it will be wrong. If the predictions are guaranteed to be wrong, will plans based on those predictions be robust to the real events and if not are there better ways of planning for uncertain futures?

Wednesday

All sessions are held in the BIS Conference Centre, 1 Victoria Street, London, SW1H 0ET

Workshop A: Communicating spatial uncetainty

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: Veronica Bowman (Dstl)

 

 

Propagating uncertainty between organisations

 

 

Veronica Bowman (Dstl)

 

 

Can interactive data visualisation help us communicate uncertainty?

 

 

Robert Grant (Kingston University)

 

 

Uncertainty: from concept to communication

 

 

Adam Chilton (Envitia)

 

 

Meteorology and dispersion – multiple sources of spatial and temporal uncertainties

 

 

Matt Hort (Met Office)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions

 



Workshop B: Uncertainty modelling for large-scale computer models

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: Philippa Spencer (Dstl)

 

 

Ensemble forecasting - a practical approach to uncertainty estimation in complex systems such as hydrometeorological forecasting

 

 

Ken Mylne (Met Office)

 

 

Calibration of WACCM's gravity waves parametrizations using spherical outputs

 

 

Kai-Lan Chang (UCL)

 

 

The mystery of the missing (probability) mass: Understanding what the IPCC actually said about climate in 2100

 

 

Erica Thompson (LSE)

 

 

A framework for parameter uncertainty in catastrophe models

 

 

Steve Jewson (Risk Management Solutions)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions

 



Workshop C: Decision making under uncertainty

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: David Lowe (Dstl)

 

 

New guidance on quality and uncertainty in government analysis – and why calculating uncertainty is not enough

 

 

Andrew Hart (FERA)

 

 

Bayesian models for cost-effectiveness analysis in the presence of structural zero costs

 

 

Gianluca Baio (UCL)

 

 

A Bayesian perspective on info-gap analysis

 

 

Ullrika Sahlin (Lund University)

 

 

Uncertainty handling in integrating decision support systems

 

 

Manuele Leonelli (University of Warwick)

 

 

Handling uncertainty using the hierarchical process model

 

 

Mike Yearworth (University of Bristol)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions

 



Workshop D: Representing uncertainty and risk in cost estimates for complex systems

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: Gary Glennon-Alty (Dstl)

 

 

Propagation patterns of cost uncertainty across the aerospace whole product life cycle

 

 

Oliver Schwabe (Rolls Royce/Cranfield University)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions

 



Workshop E: Uncertainty propagation and extrapolated model predictions

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: Martin Owen (GlaxoSmithKline)

 

 

Topological sensitivity analysis for dynamic models

 

 

Anne Babtie (Imperial)

 

 

Accelerated Stability Assessment Protocol (ASAP) studies: estimation of errors and their impact on decisions

 

 

Rachel Orr (GlaxoSmithKline)

 

 

Propagating uncertainty through models

 

 

Philip Bowtel (Procter & Gamble)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions

 



Workshop F: Public perception of risk

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: Chris Felton (Cabinet Office)

 

 

Re-thinking public responses to extreme events: understanding the relationship between risk perception, risk communication and behaviour

 

 

Brooke Rogers (King's College London)

 

 

What is `Confidence'?

 

 

Nick Hare (Aleph Insights)

 

 

Communicating risk information for decision making: the independent breast screening review

 

 

Casey Dunlop (Cancer Research UK)

 

 

TBC

 

 

Mark Gibbs (Met Office)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions

 



Workshop G: Planning for complex futures

 

 

 

 

 

09:00-11.00

        1. Talks

 

 

Chair: Simon Collander-Brown (Dstl)

 

 

Preparing for the future and dealing with uncertainty using scenario planning

 

 

James Derbyshire (Middlesex University)

 

 

11:00-11:30

Coffee

 

 

11.30-12.30

        2. Discussions