Power of ML and DL in Precision Modeling and Climate Predictions

Abstract

Machine learning (ML) and deep learning (DL) focused on models with many parameters arranged in artificial neural network architectures are increasingly being applied as powerful tools for change across various scientific disciplines, including but not limited to oceanography and climate science. In turn, these tools for “data assimilation” provide a critical path toward further harnessing the ocean's new data capacity to improve how we model and predict climate. ML algorithms have been used to solve a variety of problems related to our oceans including habitat sea level prediction, wind and wave front dressing alongside detection systems for submersible objects.ML and DL methods have also been applied FFR to improve the skill of forecasting climate. The use of these algorithms enables the extraction and interpretation of latent climate patterns, unveiling untapped potential information available in large volumes-but-complex datasets. This chapter focuses on the use and integration of Machine Learning and Deep Learning in precision modelling and climate predictions.