The Predictive Analysis of Sea Surface Temperature (SST) from the Great Barrier Reef in Australia Based on EMD-LSTM

The escalating sea surface temperature (SST) poses a significant threat to coral reef ecosystems globally, making SST prediction a pivotal research area in operational oceanography. This study compiles and analyzes SST data from the Great Barrier Reef in Australia. We explore SST prediction using the Long Short-Term Memory (LSTM) network and a hybrid Empirical Mode Decomposition (EMD)-LSTM approach, Results indicate that both models accurately forecast SST, by comparing the RMSE, MAPE, and MAE metrics, it can be concluded that the EMD-LSTM model outperforms the LSTM model in prediction effectiveness. This underscores the potential of the EMD-LSTM approach in enhancing coral reef conservation efforts by providing more precise SST forecasts, thereby mitigating the adverse impacts of temperature stress on these fragile ecosystems.