The rationale behind the success of multi-model ensembles in seasonal forecasting — I. Basic concept

By R. Hagedorn, Francisco J. Doblas-Reyes, Tim N. Palmer. 2005

720 citations recorded by the source corpus.

The DEMETER multi-model ensemble system is used to investigate the rationale behind the multi-model concept. A comprehensive documentation of the differences in the single and multi-model performance in the DEMETER hindcast data set is given. Both deterministic and probabilistic diagnostics are used and a variety of analyses demonstrate the improvements achieved by using multi-model instead of single-model ensembles. In order to understand the reason behind the multi-model superiority, basic scenarios describing how the multi-model approach can improve over singlemodel skill are discussed. It is demonstrated that multi-model superiority is caused not only by error compensation but in particular by its greater consistency and reliability.

Subfields: Atmospheric Science, Global and Planetary Change

View Paper: https://www.semanticscholar.org/paper/567cbaffe1cdbf768834832eee719108af863fb4

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

In Global and Planetary Change

Citations in the corpus

Multimodel Ensemble ENSO Prediction with CCSM and CFS (2009, result): “…presented here builds on many previous studies (i.e., Cane et al. 1986; Leetmaa and Ji 1989; Chen et al. 1995; Balmaseda et al. 1995; Ji and Leetmaa 1997; Rosati et al. 1997; Kirtman et al. 1997; Kirtman and Zebiak 1997; Stockdale 1997; Stockdale et al. 1998; Ji et al. 1994, 1996, 1998;…”
Stochastic climate theory and modeling (2015, positive): “The models can diﬀer in the dynamical core and the physical parameterizations 55,63,75 or use the same dynamical core but utilize either diﬀerent static parameters in their physical parameterizations 151 or altogether diﬀerent physics packages 6,31,106,152 .”
A verification framework for interannual-to-decadal predictions experiments (2012, background): “…still face a relatively short history in face of decadal-scale variability Finally, it is possible that gains in prediction quality may be made by multi-model ensemble approaches (e.g. Hagedorn et al. 2005), as has been the case for seasonal prediction (Robertson et al. 2004; Graham et al. 2011).”
Skill of Multimodel ENSO Probability Forecasts (2008, methodology): “Using the ECMWF model, Hagedorn et al. (2005) demonstrated that while increasing the ensemble size of a single model did increase skill, the multimodel ensemble retained a slight advantage over the single model ensemble.”
Using a Stochastic Kinetic Energy Backscatter Scheme to Improve MOGREPS Probabilistic Forecast Skill (2011, background): “This has a number of unique advantages as discussed, for example, in Park et al. (2008), Candille (2009), and Hagedorn et al. (2005).”
Skill improvement from increased ensemble size and model diversity (2014, background): “…generated in model intercomparison projects [Krishnamurti et al., 2000; Pavan and Doblas-Reyes, 2000; Barnston et al., 2003; Tebaldi et al., 2004; Hagedorn et al., 2005; DelSole, 2007; Peña and van den Dool, 2008; Casanova and Ahrens, 2009; van Oldenborgh et al., 2012; DelSole et al., 2013].”
Probabilistic Seasonal Forecasts in the North American Multimodel Ensemble: A Baseline Skill Assessment (2016, background): “A considerable advantage of multimodel ensembles (MMEs) over a single-model approach is thought to be the ability to sample forecast uncertainty due to model diversity (Palmer et al. 2004; Hagedorn et al. 2005; Doblas-Reyes et al. 2005; Kirtman and Min 2009).”
Advancing data assimilation in operational hydrologic forecasting: progresses, challenges, and emerging opportunities (2012, methodology): “The multi-model ensemble strategy is a means to address model structure uncertainty by synthesizing outcomes from multiple models representing different parameterizations of underlying physical processes and has been demonstrated to offer better predictability (Hagedorn et al., 2005).”