Deep Learning-Based Sea Surface Temperature Prediction in Korean Coastal Waters

View at publisher Summarise with VIRAT
Authors: Ji-Yeon Kim, Ki-Hwan Kim, Young-Jin Kang, Seok Chan Jeong
Cited by: 0
References: 6

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Sea surface temperature (SST) is a critical variable for monitoring ecosystems, managing fisheries resources, and analyzing climate variability. This study used NOAA OISST data to predict the next 7-day SST from the past 7 days in the Korean coastal waters and compared the prediction performance of various deep learning models. Preprocessing was applied to account for the land-sea boundary, and prediction performance was evaluated using RMSE, MAE, and MSE. Experimental results showed that the U-Net and ConvLSTM models achieved the best prediction performance, highlighting the importance of preserving spatial structure and learning spatiotemporal continuity in SST prediction.

Fields of Study & Subjects

Oceanographic and Atmospheric Processes Hydrological Forecasting Using AI Model Reduction and Neural Networks

Publication details

Journal / source
International Conference on Artificial Intelligence in Information and Communication (ICAIIC)
Pages
421–424
Publication date
24 February 2026
Language
EN