Name: Dr. Leonard Smith
Title: A Geometric View of the Ins and Outs of Simulation-based Forecasting
Time: 04:00 PM
Abstract: A Geometric View of the Ins and Outs of Simulation-based Forecasting Leonard A Smith Centre for the Analysis of Time Series, LSE Pembroke College, Oxford
Since the 1950’s weather forecasting (on every timescale) has been challenged by the nonlinearity of our models and the limits of our computational power. A geometric, nonlinear dynamics view of forecasting will be illustrated by cartoon, low-dimensional chaotic systems and an operational model (NOGAPS). In combination with consideration of a resource-allocation perspective, this view yields insight into what we can reasonably expect to be useful in future simulation-based forecasts, and what we must rule out targets. In particular, accepting model imperfection is shown to rule out some traditional forecasting goals. At the same time, doing so opens the door for many new tools and targets, with ample scope for practical application in future forecast-systems and for advancing both scientific insight and understanding. In the long run, abandoning the Perfect-Model Model (which aims for accountable probability forecasts) and taking a just-enough-decisive-information (JEDI) approach is more likely to provide genuine information regarding the future.