Using Bayesian Model Averaging to Calibrate Forecast Ensembles

By Adrian E. Raftery, Tilmann Gneiting, Fadoua Balabdaoui, Michael Polakowski. 2005

1,875 citations recorded by the source corpus.

Subfields: Atmospheric Science, Global and Planetary Change

View Paper: https://www.semanticscholar.org/paper/1ff8a921a2cba09f95a9a3dfe326d2c3ac4dbf79

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

Merging Seasonal Rainfall Forecasts from Multiple Statistical Models through Bayesian Model Averaging (2012, methodology): “BMA has been shown to outperform simple skill-based averaging in seasonal forecasting (Casanova and Ahrens 2009), and to perform well in forecasting and predictions in general (e.g., Raftery et al. 2005; Wright 2009).”
Generalized likelihood uncertainty estimation (GLUE) using adaptive Markov Chain Monte Carlo sampling (2008, background): “This is particularly true for the SCEM-UA created initial sample and suggests that averaging of predictions of different parameter combinations increases predictive capabilities, something that is commonly observed with ensemble forecasting [48,60].”
The error of representation: basic understanding (2015, positive): “(4) much like that of the algorithms in the ensemble post-processing and bias-correction literature (e.g. Glahn and Lowry, 1972; Raftery et al., 2005) in which climatological information is used to build relationships 0 1 2 3 4 5 6 −0.”
Uncertainty and Inference for Verification Measures (2007, background): “Forecasting), Hamill (1999), Raftery et al. (2005), Seaman et al. (1996), Stephenson (2000), Thornes and Stephenson (2001), and Woodcock (1976), together with some other references that are cited later in the paper.”
Treatment of uncertainty using ensemble methods: Comparison of sequential data assimilation and Bayesian model averaging (2007, methodology): “Analyses of predictive uncertainty based on a single hydrologic concept are prone to statistical bias and underestimation of uncertainty [Hoeting et al., 1999; Neuman, 2003; Raftery et al., 2003, 2005 ].”
A Bayesian approach to climate model evaluation and multi‐model averaging with an application to global mean surface temperatures from IPCC AR4 coupled climate models (2006, methodology): “This corresponds to the method of Raftery et al. [2005], who used posterior probability as a weighting”
Generating and Calibrating Probabilistic Quantitative Precipitation Forecasts from the High-Resolution NWP Model COSMO-DE (2012, methodology): “Other statistical postprocessing techniques are kernel dressing, affine kernel dressing (Bröcker and Smith 2008), and Bayesian model averaging (BMA; Raftery et al. 2005).”
Verification and Calibration of Neighborhood and Object-Based Probabilistic Precipitation Forecasts from a Multimodel Convection-Allowing Ensemble (2012, background): “…Walters 1998), spread–skill relationships (Atger 1999), fitted distribution parameters (Wilks 2002), Bayesian model averaging (Kass and Raftery 1995; Raftery et al. 2005; Sloughter et al. 2007), Gaussian ensemble dressing (GED; Roulston and Smith 2003; Wang and Bishop 2005; Fortin et al. 2006),…”
Forecast combinations: An over 50-year review (2022, background): “…weather forecast is constructed from a set of point forecasts of the same weather quantity of interest, based on perturbed initial atmospheric states (e.g., Maqsood et al., 2004; Gneiting and Raftery, 2005) and/or different model formulations (e.g., Buizza et al., 1999; Buizza et al., 2005).”
Assessing the Calibration of High-Dimensional Ensemble Forecasts Using Rank Histograms (2013, methodology): “To adjust the ensemble spread, we use the “error dressing” approach of Roulston and Smith (2003), building a new ensemble by sampling from the errors εij = yij − μij of the bias-corrected forecasts on the respective training days j = 1, .”
An integrated hydrologic Bayesian multimodel combination framework: Confronting input, parameter, and model structural uncertainty in hydrologic prediction (2006, methodology): “: ð14Þ [40] Because of the high dimensionality of this problem, it is hard and inefficient to maximize equation (14) through direct nonlinear maximization methods such as NewstonRaphson or its variants [Raftery et al., 2003].”
Probabilistic climate change predictions applying Bayesian model averaging (2007, methodology): “Raftery et al. (2005) showed the superiority of BMA in the probabilistic forecast as well as deterministic one in the case of mesoscale weather forecasts.”