An Efficient Hybrid Model to Predict Sea Surface Temperature

Abstract

The precision of sea surface temperature (SST) forecasts is essential for many uses, such as biological monitoring, maritime navigation, and climate modeling. In order to improve the accuracy of SST predictions, this study introduces a novel hybrid forecasting model that integrates Long Short-Term Memory (LSTM) networks with Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX).We overcome the drawbacks of traditional statistical methods, which often fail to capture complex non-linear relationships in oceanographic data, by using a large dataset from the National Oceanic and Atmospheric Administration (NOAA) that includes daily SST readings and relevant atmospheric variables over a significant period of time. Metrics like R-squared (R²), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) are used to carefully assess.