A transformer-based method for correcting daily SST numerical forecasting products
Authors:
Open Access
- 28 March 2025
- journal article
- research article
- Published by Frontiers Media SA in Frontiers in Earth Science
- Vol. 13, 1530475
Abstract
This study introduces applies a Transformer-based method to correct daily Sea Surface Temperature (SST) numerical forecasting products, addressing persistent challenges in short-term SST prediction. The proposed approach utilizes a Transformer model architecture to capture complex spatiotemporal dependencies in SST error fields, enabling efficient prediction of forecast errors across multiple time scales. The method was applied to SST hindcast data from the First Institute of Oceanography (FIO-COM) ocean forecasting system, focusing on the northwestern Pacific region. Results demonstrate significant improvements in forecast accuracy, with Root Mean Square Error (RMSE) reductions ranging from 38.8% for day 2 forecasts to 17.6% for day 5 forecasts. Spatial analysis reveals the method’s robust performance across diverse oceanographic regimes, including complex coastal and shelf regions where traditional models often struggle. The Transformer model showed the ability to capture and reproduce error patterns, effectively addressing both large-scale systematic biases and smaller-scale regional variations. The consistent performance across different forecast horizons suggests potential for extending the reliable forecast range of SST predictions. The findings have important implications for applications requiring precise SST forecasts, including operational oceanography, marine weather forecasting, and coupled ocean-atmosphere modeling.Keywords
This publication has 38 references indexed in Scilit:
- STA-SST: Spatio-temporal time series prediction of Moroccan Sea surface temperatureJournal of Sea Research, 2024
- Time series prediction of sea surface temperature based on BiLSTM model with attention mechanismJournal of Sea Research, 2024
- Observation-based Sea surface temperature trends in Atlantic large marine ecosystemsProgress in Oceanography, 2022
- A transformer neural network for predicting near‐surface temperatureMeteorological Applications, 2022
- Coupled data assimilation at ECMWF: current status, challenges and future developmentsQuarterly Journal of the Royal Meteorological Society, 2022
- Prediction of sea surface temperatures using deep learning neural networksSN Applied Sciences, 2020
- Numerical modelling of three-dimensional wave-current interactions in complex environment: Application to Alderney RaceApplied Ocean Research, 2020
- China published ocean forecasting system for the 21st-Century Maritime Silk Road on December 10, 2018Acta Oceanologica Sinica, 2019
- Prediction of daily sea surface temperature using efficient neural networksOcean Dynamics, 2017
- The performance of the US Navy's RELO ensemble, NCOM, HYCOM during the period of GLAD at-sea experiment in the Gulf of MexicoDeep Sea Research Part II: Topical Studies in Oceanography, 2016