Chaos & Uncertainty in Weather Forecasts
One of the most important challenges of society today is the prediction of future weather in a changing climate. We need to estimate the probability of unlikely but high-risk weather events, and the risk of droughts or floods on seasonal timescales. However, predicting the future is intrinsically uncertain and it is a great challenge to provide reliable forecasts for atmosphere and ocean which also include an estimate of the uncertainty in those predictions.
This course will provide an overview of how operational weather forecasts are generated. The students will be exposed to a real life weather forecasting situation and evaluate real forecast data from the Integrated Forecast System of the European Centre for Medium-Range Weather Forecasts, which is one of the best weather forecast models in the world. This data will be used to understand how forecasts are generated and evaluated in an operational weather centre, and the limits and challenges of today’s forecast capabilities.
The course will consist of a set of 3 lectures and 3 hands-on practical sessions. The first day will introduce the students to global weather and climate predictions systems. Day two and three will cover topics related to modelling uncertainty in forecasting systems, and verification metrics and tools used to evaluate these forecasts.
Lectures:
Chaos and uncertainty in weather forecasting (available upon request)
Uncertainty in Numerical Weather Forecasting (available upon request)
Preliminary Reading
Some interesting papers for perspective on the subject:
Slingo, J. and Palmer, T., 2011. Uncertainty in weather and climate prediction.Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 369(1956), pp.4751-4767.
Roberto Buizza, 2002. Chaos and weather prediction. ECMWF Meteorological Training Course Lecture Series
Palmer, T.N., 2000. Predicting uncertainty in forecasts of weather and climate. Reports on Progress in Physics, 63(2), p.71.
Textbooks:
Palmer, T. and Hagedorn, R. eds., 2006. Predictability of weather and climate. Cambridge University Press
Kalnay, E., 2003. Atmospheric modeling, data assimilation and predictability. Cambridge university press.
The practical sessions in the class will follow closely from the exercises conducted in an OpenIFS workshop (2015). Preliminary reading material about these exercises can be found at this link.
On course resources
The practical exercises will make use of a virtual machine instance of MetView environment. The data and scripts for the visualization exercises are provided as a part of the virtual machine instance. The exercises will follow closely from exercises described on this website, which was part of a OpenIFS training course.
Assignment
Graded assignments should be created in the assignments tool on weblearn.