Probabilistic and Feedforwardal Networks for Climate Prediction

A Bayesian Deep Learning Approach to Near-Term Climate Prediction

We consider the use of feedforward convolutional networks to predict natural variability of the sea surface temperature on the interannual timescale in the pre-industrial control simulation of the communityearth system model (cesm2).We find that a feedforward convolutional network with a densenet architecture is able to outperform a convolutional lstm in terms of predictive skill.Next, we consider a probabilistic formulation of the same network based on variational gradient descent and find that in addition to providing useful measures of predictive uncertainty, the probabilistic (bayesian) version improves on its deterministic counterpart in terms of predictive skill.Finally, we characterize the reliability of the ensemble of machine-learning models obtained in the probabilistic setting by using analysis tools developed in the context of ensemble numerical weather prediction.