Projekt

Zur Projekt-Website

High-impact extreme weather events, such as heavy rainfall or strong winds are often associated with intensive atmospheric fronts with strong gradients or deep moist convection. These events are governed by mesoscale atmospheric dynamics, operating on horizontal scales ranging from a few to several hundred kilometres. Accurate forecasts of severe weather events are of crucial relevance to society, and require the joint effort of atmospheric and mathematical scientists. This interdisciplinary project works towards an improved prediction framework. New statistical postprocessing methods are applied to correct for biases and imperfect representations of uncertainty in numerical forecasts while retaining the multivariate structure. The second phase of the project aims at the development of a next-generation comprehensive mesoscale predictive system for extreme weather and will further strengthen the connections between the two disciplines. The goal is to expand the general applicability of the statistical postprocessing methodology, to link it with new knowledge of dynamical systems and to further combine this methodology with new approaches related to extreme value theory.

• Priv.-Doz. Dr. Petra Friederichs

  Universität Bonn
  Meteorologisches Institut und Interdisziplinäres
  Zentrum für Komplexe Systeme
  Bonn
  

• Prof. Dr. Andreas Hense

  Universität Bonn
  Meteorologisches Institut
  und Interdisziplinäres Zentrum
  für Komplexe Systeme
  Bonn
  

• Prof. Dr. Martin Schlather

  Universität Mannheim
  Institut für Mathematik
  Mannheim
  

• Prof. Dr. Tilmann Gneiting

  HITS Heidelberger Institut für
  Theoretische Studien
  Computational Statistics Group
  Heidelberg
  

• Dr. Thordis L. Thorarinsdottir

  Norwegian Computing Center
  Oslo
  Norwegen
  

• Coherent evolution of potential vorticity anomalies associated with deep moist convection.

• A Mate rn based Gaussian spatial stochastic process to consistently model wind and related variables.

• A similarity-based implementation of the Schaake shuffle

• Ensemble calibration with preserved correlations: unifying and comparing ensemble copula coupling and member-by-member postprocessing.

• Conditionally Max-stable Random Fields based on log Gaussian Cox Processes