Evaluation of thermal remote sensing for detection of thermal anomalies as earthquake precursors: a case study for Malatya-Pütürge-Doganyol (Turkey) Earthquake, July 13, 2003

Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies.

Several studies in last two decades indicated that presence of positive thermal anomalies associated with seismic activities can be detected by satellite thermal sensing methods. This study evaluates the potential of thermal remote sensing for detection of thermal anomalies prior to Malatya-Pütürge-Doğanyol (Turkey) earthquake using MODIS/Terra V5 LST/E (MOD11A1) data. In the previous studies, different methods based on different approaches have been suggested. In this particular study, four of the suggested methods were selected for evaluation as well as for comparison of different approaches. The analyses were carried out for fortnight before and after the earthquake. Depending on the method 4 to 7 years of daily daytime and nighttime MOD11A1 data were utilized. Furthermore, same set of analyses carried out for non-earthquake years as well as the earthquake year for the area. The results show that when only the earthquake year considered, all the methods used for the analyses detected the LST changes successfully and consistently not only before but also after the earthquake. However, thermal anomalies were not unique for the earthquake year and were also observed in the absence of seismic activity within defined time interval. Therefore, there exist no coherent evidence that indicates a direct link between the occurrence of seismic activity and the land surface temperature anomaly for Malatya-Pütürge-Doğanyol earthquake. Based on the information extracted, it can be said that, the reason for observing LST changes even in the absence of the seismic activity is the effect of environmental factors which have considerable influence on the methods and thus the detection of LST anomalies. Therefore, it can be said that since the effect of the Sun’s irradiation is minimal during night nighttime images would be more appropriate for thermal anomaly detection purpose. The findings support the argument that not every earthquake is preceded by detectable thermal precursor (Freund 2007; Saraf et al. 2009). On the other hand, not every LST anomaly is followed by an earthquake. Additionally, since the mechanism is not very well understood yet, it is not possible to identify earthquakes which would have thermal precursor prior to the incident. Therefore, it is concluded that utilizing LST anomalies based on satellite imagery for monitoring impending earthquake would not be adequate and feasible unless the mechanism of thermal precursors are very well understood.