BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Skilful precipitation nowcasting using deep generative models of r
 adar - Suman Ravuri\, Google DeepMind
DTSTART:20220531T100000Z
DTEND:20220531T113000Z
UID:TALK174209@talks.cam.ac.uk
CONTACT:Herbie Bradley
DESCRIPTION:Precipitation nowcasting\, the high-resolution forecasting of 
 precipitation up to two hours ahead\, supports the real-world socioeconomi
 c needs of many sectors reliant on weather-dependent decision-making. Stat
 e-of-the-art operational nowcasting methods typically advect precipitation
  fields with radar-based wind estimates\, and struggle to capture importan
 t non-linear events such as convective initiations. Recently introduced de
 ep learning methods use radar to directly predict future rain rates\, free
  of physical constraints. While they accurately predict low-intensity rain
 fall\, their operational utility is limited because their lack of constrai
 nts produces blurry nowcasts at longer lead times\, yielding poor performa
 nce on rarer medium-to-heavy rain events. Here we present a deep generativ
 e model for the probabilistic nowcasting of precipitation from radar that 
 addresses these challenges. Using statistical\, economic and cognitive mea
 sures\, we show that our method provides improved forecast quality\, forec
 ast consistency and forecast value. Our model produces realistic and spati
 otemporally consistent predictions over regions up to 1\,536 km × 1
 \,280 km and with lead times from 5–90 min ahead. Using a systematic
  evaluation by more than 50 expert meteorologists\, we show that our gener
 ative model ranked first for its accuracy and usefulness in 89% of cases a
 gainst two competitive methods. When verified quantitatively\, these nowca
 sts are skillful without resorting to blurring. We show that generative no
 wcasting can provide probabilistic predictions that improve forecast value
  and support operational utility\, and at resolutions and lead times where
  alternative methods struggle.
LOCATION:Bullard Lecture Theatre\, in the Wolfson Building at Madingley Ri
 se
END:VEVENT
END:VCALENDAR