%0 Journal Article %1 Siegert2016Bayesian %A Siegert, Stefan %A Stephenson, David B. %A Sansom, Philip G. %A Scaife, Adam A. %A Eade, Rosie %A Arribas, Alberto %D 2016 %I American Meteorological Society %J Journal of Climate %K euporias %N 3 %P 995--1012 %R 10.1175/jcli-d-15-0196.1 %T A Bayesian Framework for Verification and Recalibration of Ensemble Forecasts: How Uncertain is NAO Predictability? %U http://dx.doi.org/10.1175/jcli-d-15-0196.1 %V 29 %X AbstractPredictability estimates of ensemble prediction systems are uncertain because of limited numbers of past forecasts and observations. To account for such uncertainty, this paper proposes a Bayesian inferential framework that provides a simple 6-parameter representation of ensemble forecasting systems and the corresponding observations. The framework is probabilistic and thus allows for quantifying uncertainty in predictability measures, such as correlation skill and signal-to-noise ratios. It also provides a natural way to produce recalibrated probabilistic predictions from uncalibrated ensembles forecasts.The framework is used to address important questions concerning the skill of winter hindcasts of the North Atlantic Oscillation for 1992?2011 issued by the Met Office Global Seasonal Forecast System, version 5 (GloSea5), climate prediction system. Although there is much uncertainty in the correlation between ensemble mean and observations, there is strong evidence of skill: the 95\% credible interval of the correlation coefficient of 0.19, 0.68 does not overlap zero. There is also strong evidence that the forecasts are not exchangeable with the observations: with over 99\% certainty, the signal-to-noise ratio of the forecasts is smaller than the signal-to-noise ratio of the observations, which suggests that raw forecasts should not be taken as representative scenarios of the observations. Forecast recalibration is thus required, which can be coherently addressed within the proposed framework. AbstractPredictability estimates of ensemble prediction systems are uncertain because of limited numbers of past forecasts and observations. To account for such uncertainty, this paper proposes a Bayesian inferential framework that provides a simple 6-parameter representation of ensemble forecasting systems and the corresponding observations. The framework is probabilistic and thus allows for quantifying uncertainty in predictability measures, such as correlation skill and signal-to-noise ratios. It also provides a natural way to produce recalibrated probabilistic predictions from uncalibrated ensembles forecasts.The framework is used to address important questions concerning the skill of winter hindcasts of the North Atlantic Oscillation for 1992?2011 issued by the Met Office Global Seasonal Forecast System, version 5 (GloSea5), climate prediction system. Although there is much uncertainty in the correlation between ensemble mean and observations, there is strong evidence of skill: the 95\% credible interval of the correlation coefficient of 0.19, 0.68 does not overlap zero. There is also strong evidence that the forecasts are not exchangeable with the observations: with over 99\% certainty, the signal-to-noise ratio of the forecasts is smaller than the signal-to-noise ratio of the observations, which suggests that raw forecasts should not be taken as representative scenarios of the observations. Forecast recalibration is thus required, which can be coherently addressed within the proposed framework.

@article{Siegert2016Bayesian, abstract = {{AbstractPredictability} estimates of ensemble prediction systems are uncertain because of limited numbers of past forecasts and observations. To account for such uncertainty, this paper proposes a Bayesian inferential framework that provides a simple 6-parameter representation of ensemble forecasting systems and the corresponding observations. The framework is probabilistic and thus allows for quantifying uncertainty in predictability measures, such as correlation skill and signal-to-noise ratios. It also provides a natural way to produce recalibrated probabilistic predictions from uncalibrated ensembles {forecasts.The} framework is used to address important questions concerning the skill of winter hindcasts of the North Atlantic Oscillation for 1992?2011 issued by the Met Office Global Seasonal Forecast System, version 5 ({GloSea5}), climate prediction system. Although there is much uncertainty in the correlation between ensemble mean and observations, there is strong evidence of skill: the 95\% credible interval of the correlation coefficient of [0.19, 0.68] does not overlap zero. There is also strong evidence that the forecasts are not exchangeable with the observations: with over 99\% certainty, the signal-to-noise ratio of the forecasts is smaller than the signal-to-noise ratio of the observations, which suggests that raw forecasts should not be taken as representative scenarios of the observations. Forecast recalibration is thus required, which can be coherently addressed within the proposed framework. {AbstractPredictability} estimates of ensemble prediction systems are uncertain because of limited numbers of past forecasts and observations. To account for such uncertainty, this paper proposes a Bayesian inferential framework that provides a simple 6-parameter representation of ensemble forecasting systems and the corresponding observations. The framework is probabilistic and thus allows for quantifying uncertainty in predictability measures, such as correlation skill and signal-to-noise ratios. It also provides a natural way to produce recalibrated probabilistic predictions from uncalibrated ensembles {forecasts.The} framework is used to address important questions concerning the skill of winter hindcasts of the North Atlantic Oscillation for 1992?2011 issued by the Met Office Global Seasonal Forecast System, version 5 ({GloSea5}), climate prediction system. Although there is much uncertainty in the correlation between ensemble mean and observations, there is strong evidence of skill: the 95\% credible interval of the correlation coefficient of [0.19, 0.68] does not overlap zero. There is also strong evidence that the forecasts are not exchangeable with the observations: with over 99\% certainty, the signal-to-noise ratio of the forecasts is smaller than the signal-to-noise ratio of the observations, which suggests that raw forecasts should not be taken as representative scenarios of the observations. Forecast recalibration is thus required, which can be coherently addressed within the proposed framework.}, added-at = {2019-08-14T18:35:35.000+0200}, author = {Siegert, Stefan and Stephenson, David B. and Sansom, Philip G. and Scaife, Adam A. and Eade, Rosie and Arribas, Alberto}, biburl = {https://www.bibsonomy.org/bibtex/264f51cee85ccbd3746903aed209d10ff/karinawilliams}, citeulike-article-id = {13798442}, citeulike-linkout-0 = {http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-15-0196.1}, citeulike-linkout-1 = {http://dx.doi.org/10.1175/jcli-d-15-0196.1}, day = 8, doi = {10.1175/jcli-d-15-0196.1}, interhash = {621d3c6818d19881a1ae70c8523b5892}, intrahash = {64f51cee85ccbd3746903aed209d10ff}, issn = {0894-8755}, journal = {Journal of Climate}, keywords = {euporias}, month = feb, number = 3, pages = {995--1012}, posted-at = {2015-10-12 10:04:01}, priority = {2}, publisher = {American Meteorological Society}, timestamp = {2019-08-14T18:35:35.000+0200}, title = {A Bayesian Framework for Verification and Recalibration of Ensemble Forecasts: How Uncertain is {NAO} Predictability?}, url = {http://dx.doi.org/10.1175/jcli-d-15-0196.1}, volume = 29, year = 2016 }