Accepted author manuscript
Final published version
Research output: Contribution to Journal/Magazine › Conference article › peer-review
Research output: Contribution to Journal/Magazine › Conference article › peer-review
}
TY - JOUR
T1 - Ensembling geophysical models with Bayesian Neural Networks
AU - Sengupta, Ushnish
AU - Amos, Matt
AU - Hosking, J. S.
AU - Rasmussen, Carl Edward
AU - Juniper, Matthew
AU - Young, Paul
PY - 2020/10/7
Y1 - 2020/10/7
N2 - Ensembles of geophysical models improve projection accuracy and express uncertainties. We develop a novel data-driven ensembling strategy for combining geophysical models using Bayesian Neural Networks, which infers spatiotemporally varying model weights and bias while accounting for heteroscedastic uncertainties in the observations. This produces more accurate and uncertainty-aware projections without sacrificing interpretability. Applied to the prediction of total column ozone from an ensemble of 15 chemistry-climate models, we find that the Bayesian neural network ensemble (BayNNE) outperforms existing ensembling methods, achieving a 49.4% reduction in RMSE for temporal extrapolation, and a 67.4% reduction in RMSE for polar data voids, compared to a weighted mean. Uncertainty is also well-characterized, with 90.6% of the data points in our extrapolation validation dataset lying within 2 standard deviations and 98.5% within 3 standard deviations.
AB - Ensembles of geophysical models improve projection accuracy and express uncertainties. We develop a novel data-driven ensembling strategy for combining geophysical models using Bayesian Neural Networks, which infers spatiotemporally varying model weights and bias while accounting for heteroscedastic uncertainties in the observations. This produces more accurate and uncertainty-aware projections without sacrificing interpretability. Applied to the prediction of total column ozone from an ensemble of 15 chemistry-climate models, we find that the Bayesian neural network ensemble (BayNNE) outperforms existing ensembling methods, achieving a 49.4% reduction in RMSE for temporal extrapolation, and a 67.4% reduction in RMSE for polar data voids, compared to a weighted mean. Uncertainty is also well-characterized, with 90.6% of the data points in our extrapolation validation dataset lying within 2 standard deviations and 98.5% within 3 standard deviations.
M3 - Conference article
VL - 33
SP - 1
EP - 13
JO - Advances in Neural Information Processing Systems
JF - Advances in Neural Information Processing Systems
ER -