dot
Detailansicht
Katalogkarte GBA
Katalogkarte ISBD
Suche präzisieren
Drucken
Download RIS
Hier klicken, um den Treffer aus der Auswahl zu entfernen
Titel How can model error characteristics be utilised to improve climate projections?
VerfasserIn Andreas Gobiet, Martin Suklitsch, Georg Heinrich
Konferenz EGU General Assembly 2013
Medientyp Artikel
Sprache Englisch
Digitales Dokument PDF
Erschienen In: GRA - Volume 15 (2013)
Datensatznummer 250080077
 
Zusammenfassung
The discussion on how to improve climate projections of a multi-model ensemble by exploiting information on model errors has a long tradition. Most studies follow the strategy to weight simulations according to their error characteristics or in a Bayesian framework. However, the potential to find optimal weights is currently very limited and none of the weighting approaches gained widespread acceptance so far. In this study, an alternative strategy is pursued: The systematic effect of a particular type of model error („intensity-dependent error“) on the expected climate change signal of a multi-model ensemble and its spread is utilised. A theory of the influence of intensity-dependent model errors on the climate change signal is developed and it is investigated, whether expected climate change is biased and uncertainty is inflated due to such errors. In addition, empirical correction methods are analysed in order to mitigate potential biases in expected climate change and to constrain uncertainty. The theory is applied to the ENSEMBLES regional multi-model ensemble. The results indicate that expected temperature change at the end of the 21st century from the ENSEMBLES ensemble is regionally biased with an over-estimation of up to +0.5 K in Eastern Europe and France in summer and an underestimation of about -0.4 K in Scandinavia. In addition, it is shown that intensity-dependent errors tend to increase the spread of a multi-model ensemble, which indicates the potential to reduce uncertainty in climate projections by a suitable correction.