Artificial Neural Network based thermal anomalies associated with earthquakes in Pakistan from MODIS LST

There are several reports about thermal anomalies associated with the impending earthquakes (EQs); among which underlying mechanism is linked main shock long, intermediate and short-term precursors. In this paper, we analyzed from Moderate Resolution Imaging Spectroradiometer (MODIS) based Land Surface Temperature (LST) of three different magnitudes shallow depth EQs in Pakistan. Our focus to investigate by statistical approach Artificial Neural Network (ANN) spatial temporal LST values within 10 days before after as After implementing ANN approach, revealed that occurred 1–10 M w > 6.0 EQs. However, a low intensity anomaly also recorded 20–25 5.2 EQ. We study magnitude dependent it more likely occur EQ before/after day. The results depict not clearly shock, locates outside window shock. Similarly, two out events caused post-seismic less compared pre-seismic anomalies. high large than post seismic all suggesting be reliable precursor short interval an • MODIS time series trained through Network. Thermal investigated earthquakes. Significant pre observed. Two exhibit