제목: Deep learning for climate prediction and projection​​​​

In first part of this talk, I will demonstrate that deep learning successfully detects the emerging climate change signals in daily precipitation fields during the observed record. Accordingly, we trained a convolutional neural network with daily precipitation fields and annual mean global mean surface air temperature data obtained from an ensemble of present-day and future climate model simulations. After applying the algorithm to the observational record, we found that the daily precipitation data represented an excellent predictor for the observed planetary warming, as they have showed a clear deviation from natural variability since the mid-2010s. In second part, a deep-learning-based global oceanic data assimilation system, will be referred as DeepDA system, will be introduced. DeepDA is formulated by incorporating a partial convolutional neural network and a generative adversarial network (GAN). The partial convolution acts as an observation operator that projects the irregular observational information on gridded fields, and the GAN model brings in the observational information from previous time frames. Observing system simulation experiments showed that the analysis error in the DeepDA-produced three-dimensional temperature is systematically reduced compared to both the background and observed values.