BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY: Ensembling climate models with Gaussian processes to better chara
 cterise future extremes -   Kenza Tazi (Department of Engineering\, Univer
 sity of Cambridge and British Antarctic Survey)
DTSTART:20241008T110000Z
DTEND:20241008T120000Z
UID:TALK221146@talks.cam.ac.uk
CONTACT:Vichawan (Print) Sakulsupich
DESCRIPTION:\nRegional climate models (RCMs) are useful tools used by many
  scientists and policymakers to represent possible climate futures. Howeve
 r\, predictions from these models are far from perfect. RCMs suffer not on
 ly from the uncertainty of different emission scenarios and inter and intr
 a-model variability\, but also from regionally specific challenges such as
  complex topography\, limited direct measurements for calibration\, and un
 suitable parametrisations schemes. Future predictions can therefore be con
 tradictory and lack consensus. Furthermore\, they are generally simply ave
 raged with the model spread as a proxy for uncertainty. In this talk\, I w
 ill present a probabilistic ensembling method to combine different RCM out
 puts based on Gaussian process regression. The aim is to quantify climate 
 extremes in a more principled way using the probability distributions of p
 ast observations. More specifically\, we’ll look at precipitation over H
 igh Mountain Asia\, an area providing water resources to approximately 2 b
 illion people.\n\n----------\n\nCAS UnionRoad is inviting you to a schedul
 ed Zoom meeting.\n\nTopic: CAS seminar: Kenza Tazi\nTime: Oct 8\, 2024 12:
 00 PM London\n\nJoin Zoom Meeting\nhttps://us02web.zoom.us/j/86392541597?p
 wd=uagQkzgbCKXTli6ayLR8gjMMufQZpo.1\n\nMeeting ID: 863 9254 1597\nPasscode
 : 830952\n\n
LOCATION:Chemistry Dept\, Unilever Lecture Theatre and Zoom
END:VEVENT
END:VCALENDAR