Prediction of the earthquake magnitude by time series methods along the East Anatolian Fault, Turkey

In this study, the magnitude of an earthquake in the East Anatolian Fault (EAF) of Turkey are predicted based on previous earthquakes whose magnitudes are four or more by two-time series methods, namely autoregressive integrated moving average (ARIMA) and singular spectrum analysis (SSA). These methods are quite new in seismology despite being successful techniques in other branches of science. We use ARIMA and SSA models to train and predict the mean and maximum values of the earthquakes' magnitudes due to seismological events between years 1900 and 2019. 447 earthquake magnitudes between 1900 and 1995 are used for training models, and then 447 magnitudes between 1995 and 2019 are taken into account for testing. The root mean square error (RMSE) is calculated to evaluate the accuracy of each model. The results demonstrate that the SSA model is better than the ARIMA model to predict the earthquake magnitude. Hence, for the years 2020 to 2030 the magnitude of an earthquake is forecasted using the SSA model. The result shows that the highest magnitude of earthquake is forecasted for the year 2021 in magnitude level of 4.0–5.9.