Investigating the relationship between earthquake occurrences and climate change using RNN-based deep learning approach

Earthquakes have been the most well-known and commonly occurring geophysical disaster on Earth. The causes of their occurrences have been deduced by the exploration of various geological phenomena. Global temperature is a major climate variable that has a wide impact on the Earth’s ecology and civilization. Global temperatures have also progressively risen as a result of the industrial revolution and urbanization. The correlation between earthquakes and rising global temperatures has been proposed before but has not been empirically verified. An investigation has been made to find the correlation between the two dynamics using the most popular deep learning-based tools. The magnitude of globally occurring earthquakes and global temperature fluctuations are used as the experimental dataset for this study. The dataset has been fed into neural networks with four types of recurrent units: simple recurrent neural networks, gated recurrent units, long short-term memory units, and bidirectional long short memory cells. These models have been validated using well-known metrics, namely, mean absolute error (MAE), mean squared error (MSE), mean absolute percentage error (MAPE), and log-cosh loss. The value of these cost functions converges to a small value for the long short-term memory model (MAE = 0.31, MSE = 0.19, MAPE = 0.63, and log-cosh loss = 0.14), which signifies the existence of a strong correlation between the magnitude of earthquakes and global temperature fluctuations.