BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Machine learning seminar series - Machine-learning-model-data-inte
 gration for a better understanding of the Earth System
DTSTART:20201124T160000Z
DTEND:20201124T170000Z
DTSTAMP:20260722T142300Z
UID:indico-event-227@events.ecmwf.int
CONTACT:events@ecmwf.int
DESCRIPTION:\n\n\n\n\n\nHost\n\nPeter Deuben (ECMWF)\n\nSpeaker\n\nMarkus 
 Reichstein\, born 1972 in Kiel\, Germany\, studied Landscape Ecology at th
 e University of Münster\; 1998-2003 research assistant at the University 
 of Bayreuth (PhD in 2001)\, EU-Marie-Curie research fellow at the Universi
 ty of Tuscia (Italy\, with research stays at the University of Montana and
  the University of California\, Berkeley\, USA) 2004 to 2006\; from 2006 t
 o 2012 Research Group Leader at the Max Planck Institute of Biogeochemistr
 y\, Jena.\n\nDirector of the Department of Biogeochemical Integration at t
 he Max Planck Institute of Biogeochemistry in Jena since 2012 and Professo
 r of Global Geoecology at the Friedrich Schiller University Jena since 201
 4. His research interests include data-driven Earth system science\, the a
 pplication of artificial intelligence/machine learning\, global biogeochem
 ical cycles\, soils in the Earth system\, and climate extremes and system 
 resilience. In 2020 Markus Reichstein was awarded the Gottfried Wilhelm Le
 ibniz Prize\, in 2019 the ERC Synergy Grant USMILE. He is the 2018 award w
 inner for the Piers J. Sellers Mid-Career Award of the American Geophysica
 l Union (AGU) and received the Max Planck Research Award of the Alexander 
 von Humboldt Foundation and Max Planck Society in 2013.\n\nAbstract\n\nThe
  Earth is a complex dynamic networked system. Machine learning\, i.e. deri
 vation of computational models from data\, has already made important cont
 ributions to predict and understand components of the Earth system\, speci
 fically in climate\, remote sensing and environmental sciences. For instan
 ce\, classifications of land cover types\, prediction of land-atmosphere a
 nd ocean-atmosphere exchange\, or detection of extreme events have greatly
  benefited from these approaches. Such data-driven information has already
  changed how Earth system models are evaluated and further developed. Howe
 ver\, many studies have not yet sufficiently addressed and exploited dynam
 ic aspects of systems\, such as memory effects for prediction and effects 
 of spatial context\, e.g. for classification and change detection. In part
 icular new developments in deep learning offer great potential to overcome
  these limitations.\n\nYet\, a key challenge and opportunity is to integra
 te (physical-biological) system modeling approaches with machine learning 
 into hybrid modeling approaches\, which combines physical consistency and 
 machine learning versatility. A couple of examples are given with focus on
  the terrestrial biosphere\, where the combination of system-based and mac
 hine-learning-based modelling helps our understanding of aspects of the Ea
 rth system.\n\n\n\n\n\nhttps://events.ecmwf.int/event/227/
IMAGE;VALUE=URI:https://events.ecmwf.int/event/227/logo-528950275.png
LOCATION:16:00 GMT
URL:https://events.ecmwf.int/event/227/
END:VEVENT
END:VCALENDAR