Estimation of Major Seismic Activities in Thailand Using Fourier Transform

A typical earthquake dataset's time series was never regularly sampled. Nonetheless, the standard Fast Fourier Transform (FFT) solely assumes uniformly spaced samples. In this paper, we handle this mathematical challenge by applying a modified version of a conventional FFT called the Non-Uniform Fast Fourier Transform (NFFT) to analyze the seismic data in Thailand. Eight major fault lines are analyzed to determine whether there have been any changes in seismic cycle duration or intensity. Later, the cycle length is utilized to estimate the upcoming significant earthquake occurrences. Our estimation has an average absolute error of 0.13 years. We developed a web-based application that allows users to select the region and presents the analytical results on an open-source map-oriented user interface, facilitating more precise assessments and preparedness actions for potential seismic risks in Thailand.