Naslov
Model evaluation and ensemble modelling of surface-level ozone in Europe and North America in the context of AQMEII

Autori
Efisio, Solazzo ; Roberto, Bianconi ; Robert, Vautard ; K. Wyat, Appel ; Bertrand, Bessagnet ; Jørgen, Brandt ; Jesper H., Christensen ; Charles, Chemel ; Isabelle, Coll, Hugo, Denier van der Gon ; Joana, Ferreira ; Renate, Forkel ; Xavier, V. Francis ; George, Grell ; Paola, Grossi ; Ayoe B., Hansen ; Jeričević, Amela ; Lukša, Kraljević ; Ana Isabel, Miranda ; Michael D., Moran ; Uarporn, Nopmongco ; Guido, Pirovano ; Marje, Prank ; Angelo, Riccio ; Karine N., Sartelet ; Martijn, Schaap ; Jeremy D., Silver ; Ranjeet S., Sokhi ; Julius, Vira ; Johannes, Werhahn ; Ralf, Wolke ; Greg, Yarwood ; Junhua, Zhang ; S.Trivikrama, Rao ; Stefano, Galmarini

Izvornik
Atmospheric environment (1994) (1352-2310) 53 (2012), SI; 60-74

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
AQMEII; Clustering; Error minimisation; Multi-model ensemble; Ozone

Sažetak
More than ten state-of-the-art regional air quality models have been applied as part of the Air Quality Model Evaluation International Initiative (AQMEII). These models were run by twenty independent groups in Europe and North America. Standardised modelling outputs, over a full year (2006), from each group have been shared on the web distributed ENSEMBLE system, which allows for statistical and ensemble analyses to be performed. The simulations of ground-level ozone concentrations issued from the models are collectively examined in an ensemble fashion, and evaluated with a large set of observations in both continents. The scale of the exercise is unprecedented and offers a unique opportunity to investigate methodologies for generating skilful ensembles of models. Despite the remarkable progress of ensemble air quality modelling over the past decade, there still are outstanding questions regarding this technique. Among them, what is the best and most beneficial way to build an ensemble of members? How to determine the optimum size of the ensemble in order to capture data variability as well as keeping the error low? We try to address these questions by looking at optimal ensemble size and quality of the members. The analysis carried out is based on systematic minimization of the model error and it is of direct relevance for diagnostic/probabilistic model evaluation. We show that the most commonly used multi-model approach, namely the average over all available members, can be outperformed by subsets of members optimally selected in terms of bias, error, and correlation. More importantly, this result does not strictly depend on the skills of the individual members, but requires the inclusion of low ranking-skill members. We apply a methodology to discern among members and to build a skilful ensemble based on model association and data clustering, which makes no use of priori knowledge of model skill. Results show that while the methodology needs further refinements, by optimally selecting the cluster distance and association criteria, this approach can be useful for model applications beyond those strictly related to model evaluation, such as air quality forecasting.

Izvorni jezik
Engleski

Znanstvena područja
Fizika, Geologija, Kemija

Napomena
Rad je prezentiran na skupu An International Initiative for the Evaluation of Regional-Scale Air Quality Models - Phase 1 (AQMEII) ; S. Trivikrama Rao, S. Galmarini, Douw G. Steyn (ur.).

POVEZANOST RADA