GREENCYCLES II

TW2 - Model benchmarking and data assimilation



Event Dates: 13th-16th September 2011
Location: Jena, Germany
Organisers: S. Zaehle, Max Planck Institute for Biogeochemistry, Germany
ECTS Point Allocation: 3

From 13-16th September 2011, 15 Greencyclists and 7 external participants participated in the Greencycles training workshop 2 (model benchmarking and data assimilation). The course covered a range of topics from model evaluation to model sensitivity studies to data assimilation techniques.

The first day was dedicated to the introduction to ecosystem modelling and model evaluation techiques with lectures by Sönke Zaehle (modelling), Greencycles fellow Daniela Dalmonech (T1.1, model evaluation techniques), Miguel Mahecha (MPI Biogeochemistry; analysis of model behaviour in the time-frequency domain) and Greencycles fellow Jana Koloassa (T1.5: comparison to remote-sensing data). The lectures were then followed by a hands-on tutorial to put the learnt content into practice using data and model output from "real" model evaluation projects under the supervision of Daniela Dalmonech and Sönke Zaehle.

The second day continued with lectures on model evaluation (treating errors in data, Gitta Lasslop, MPI Meteorology; and remote-sensing data, Martin Jung MPI Biogeochemistry). We then proceeded to part two of the course with an introduction to model sensitivity studies and useful assessment techniques by Gregor Schürmann (MPI Biogeochemistry), which was followed by a hands-on tutorial using a simple carbon-cycle model under the supervision of Gregor Schürmann and Sönke Zaehle.

three provided an introduction to data assimilation (Sönke Zaehle), followed by a detailed treatment of gradient-based assimilation methods (Thomas Kaminski, FastOpt), supplemented by a case-study for such a system (Marko Scholze, Univ. Bristol). The latter was then subject also of a hands-on tutorial on a simple carbon-cycle data assimilation system under the supervision of Marko Scholze. The day then culminated in a joint dinner at the Papiermühle in downtown Jena, where lecturers and students were able to converse accompanied by traditional Thuringian food and beer.

Day four then provided an overview on further model-data integration techniques with lectures on monte-carlo methods (Jens Kattge) and gradient-based search algorithms (Nuno Carvalhais, both MPI Biogeochemistry), as well as Ensemble Kalman filters (Sönke Zaehle). The course ended then in a general discussion on the possibilities to apply the learnt material on the Greencycles fellows’ projects.


Participants

  1. Callum Berridge, Vrije Universiteit, Netherlands – GC
  2. Ioannis Bistinas, Technical University of Lisbon, Portugal – GC
  3. Maarten Braakhekke, Max Planck Institute for Biogeochemistry, Germany
  4. Chao Ting Chang, CREAF, Spain – GC
  5. Katherine Crichton, LGGE, France – GC
  6. Altug Ekici, Max Planck Institute for Biogeochemistry, Germany – GC
  7. Aideen Foley, University of Cambridge, Uk – GC
  8. Matthias Forkel, Max Planck Institute for Biogeochemistry, Germany
  9. László Hunor Hajdu, University of Cambridge, UK – GC
  10. Jana Kolassa, Estellus, France – GC
  11. Roberto Kretschmer, Max Planck Institute for Biogeochemistry, Germany
  12. Yogesh Kumkar, Max Planck Institute for Biogeochemistry, Germany
  13. Gerardo López Saldaña, Technical University of Lisbon, Portugal – GC
  14. Alex Marti-Donati, University of Exeter, UK – GC
  15. Maria Martin Calvo, Imperial college London, UK – GC
  16. Catherine Morfopoulos, Imperial College London, UK – GC
  17. Tonatiuh Guillermo Nuez Ramirez, Max Planck Institute for Biogeochemistry, Germany
  18. Dominik Sperlich, CREAF, Spain – GC
  19. Martin Thurner, Max Planck Institute for Biogeochemistry, Germany
  20. Ulrich Weber, Max Planck Institute for Biogeochemistry, Germany
  21. Matteo Willeit, Potsdam Institute for Climate Impact Research, Germany – GC
  22. Rozenn Keribin, University of Cambridge, UK – GC