BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Representing Model Uncertainty in the ECMWF Convection Scheme - Dr
  Hannah Arnold (University of Oxford)
DTSTART:20140210T141500Z
DTEND:20140210T153000Z
UID:TALK49280@talks.cam.ac.uk
CONTACT:Dr Amanda Maycock
DESCRIPTION:Probabilistic forecasts enable users to make better informed d
 ecisions. In particular\, it is important that these probabilistic forecas
 ts are reliable\, i.e. the forecast probability of an event matches the su
 bsequent observed probability of that event taking place. In order to prod
 uce reliable probabilistic weather forecasts\, it is important to account 
 for all sources of error in atmospheric models. In the case of weather pre
 diction\, the two main sources of error are due to initial condition uncer
 tainty and model uncertainty – this talk will focus on how to represent 
 the latter. Two approaches are considered. A perturbed parameter approach 
 identifies uncertain parameters in the physical parametrisation schemes an
 d varies their values between forecasts. Stochastic parametrisation scheme
 s are also considered\, which introduce random numbers into the forecast e
 quations to represent the effect of errors in the forecast model on the ev
 olution of the forecast. These two techniques will be illustrated in terms
  of uncertainty due to the parametrisation of convection\, and tested usin
 g the European Centre for Medium Range Weather Forecasts global NWP model.
 \n
LOCATION:Unilever Lecture Theatre\, Department of Chemistry
END:VEVENT
END:VCALENDAR