Assessing the potential of state-of-the-art machine learning and physics-informed machine learning in predicting sea surface temperature

TL;DR

This paper evaluates the effectiveness of traditional machine learning and physics-informed machine learning models in predicting sea surface temperature using atmospheric data, highlighting their potential as alternatives to numerical methods.

Contribution

It benchmarks ML and PINN models for SST prediction, demonstrating their predictive capabilities with real atmospheric and oceanic data.

Findings

ML and PINN models show promising accuracy in SST forecasting.

Physics-informed models incorporate physical laws, improving interpretability.

Results suggest potential for real-time ocean monitoring applications.

Abstract

The growing adoption of machine learning (ML) in modelling atmospheric and oceanic processes offers a promising alternative to traditional numerical methods. It is essential to benchmark the performance of both ML and physics-informed ML (PINN) models to evaluate their predictive skill, particularly for short- to medium-term forecasting. In this study, we utilize gridded sea surface temperature (SST) data and six atmospheric predictors (cloud cover, relative humidity, solar radiation, surface pressure, u-component of velocity, and v-component of velocity) to capture both spatial and temporal patterns in SST predictions.