Objective Discrimination and Pooling Models in the Ensemble
Data assimilating and pooling the model predictions in the multi-model ensemble, described in this paper, are based on the techniques of approximation and regularization of multidimensional vectors in the linear Euclidean space with the use of the non-orthogonal vector basis. This approach has been successfully applied to (i) the time series of the annual temperatures averaged over the globe and northern hemisphere corresponding to the last 100years, and (ii) the fields of concentrations of atmospheric pollutants over Europe. Quantitative estimates of the efficiency of the proposed technique are presented in the paper.
KeywordsRoot Mean Square Error Numerical Weather Prediction Nitrogen Dioxide Objective Discrimination Observational Vector
This work was partially funded by the Russian Foundation for Basic Research (grant # 08-05-00569-a). We would like to acknowledge the participants of the project GEMS for kindly providing all necessary data sets.
- 1.Genikhovich EL, Pavlova TV, Kattsov VM (2010) On combining climate models in ensembles. Proc Main Geophys Observatory 561:28–46 (in Russian)Google Scholar
- 3.Mallet V, Sportisse B (2006) Ensemble-based air quality forecasts: a multimodel approach applied to ozone. J Geophys Res 111:D18302.1–D18302.11. doi:10.1029/2005JD006675Google Scholar
- 4.Mallet V, Stoltz G, Mauricette B (2009) Ozone ensemble forecast with machine learning algorithms. J Geophys Res 114:D05307.1–D05302.13Google Scholar
- 6.Met Office Hadley Centre observations datasets. http://www.cru.uea.ac.uk/cru/data/temperature/; http://hadobs.metoffice.com/hadcrut3/diagnostics/global