BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Using deep learning for precipitation nowcasting applications - Cl
 aire Bartholomew | Met Office
DTSTART:20210525T100000Z
DTEND:20210525T113000Z
UID:TALK159061@talks.cam.ac.uk
CONTACT:87364
DESCRIPTION:The UK has a very good radar network coverage\, with a continu
 ally growing amount of available precipitation observation data. Machine l
 earning offers an opportunity to harness more of the potential of this val
 uable data\, especially for nowcasting applications (0-2 hour forecasts). 
 The London Terminal Manoeuvring Area is (normally!) one of the busiest sec
 tors of airspace in the world\, where improved short-term forecasts would 
 be particularly beneficial for air traffic management. The aim of this wor
 k is to understand if and how deep learning methods could add value over m
 ore traditional precipitation nowcasting methods.\nResults from a neural n
 etwork that makes use of latent space to represent the uncertainty in the 
 system will be presented. Objective verification results and example case 
 studies will be shown\, comparing the machine learning model predictions t
 o the Met Office’s current operational model. An ensemble approach using
  this model will be presented as well as consideration of challenges in th
 ese methods.\n\nAt the beginning of the talk I will also give a brief over
 view of other Data Science activities going on at the Met Office. 
LOCATION:https://zoom.us/j/6708259482?pwd=Qk03U3hxZWNJZUZpT2pVZnFtU2RRUT09
END:VEVENT
END:VCALENDAR