Publications

Hafeez, A; Ehsan, M; Abbas, A; Shah, M; Shahzad, R (2022). Machine learning-based thermal anomalies detection from MODIS LST associated with the M-w 7.7 Awaran, Pakistan earthquake. NATURAL HAZARDS, 111(2), 2097-2115.

Abstract
Satellite based thermal anomaly occurs as a substantial precursor for strong earthquakes, as the need for earthquake precursor detection has very important for impending main shock estimation. In this study, Land Surface Temperature (LST) for both day- and night-time from Moderate Resolution Imaging Spectroradiometer (MODIS) satellite have been analyzed to monitor anomalous variation before and after the Awaran Pakistan earthquake on September 24, 2013 (M-w 7.7). We aim to find a common monitoring time window of pre-and post-seismic LST irregularities by different detecting techniques (e.g., Inter Quartile Range (IQR), wavelet transformation, Auto Regressive Integrated Moving Average (ARIMA), and Neural Network (NN)). For this purpose, three months before and three months subsequent to main shock data are analyzed for Awaran earthquake. Interestingly, every method shows an irregular variation of LST within 4-7 days before the main shock. Similarly, this analysis also pointed out a rise in temperature within 2-4 days after the main shock as post-earthquake responses. This shows the capability of LST anomalies for possible earthquake anomalies and the importance of Machine Learning (ML) techniques for detecting earthquake anomalies to support lithosphere-atmosphere-ionosphere coupling (LAIC) hypothesis for future studies.

DOI:
10.1007/s11069-021-05131-8

ISSN:
1573-0840