Skip to yearly menu bar Skip to main content

Workshop: Time Series Workshop

Morning Poster Session: Modeling the El Niño Southern Oscillation with Neural Differential Equations

Ludovico Giorgini


We use a Neural Ordinary Differential Equation to model and predict the seasonal to interannual variability of El Niño Southern Oscillation (ENSO). We train our neural network model using partial observations involving only sea surface temperature data. Our approach is computationally inexpensive, it reproduces the main seasonal features of ENSO, and exhibits robust predictions skills.