Long-term statistical evidence proving the correspondence between tir anomalies and earthquakes is still absent.

Recently, researchers have conducted long-term statistical studies to prove the correspondence between earthquakes and thermal infrared (TIR) anomalies. Upon obtaining relatively high true positive rates (TPR), it was concluded that TIR anomalies are closely related to earthquakes. However, the temporal-spatial clustering of earthquakes and overly large temporal-spatial windows will also contribute to high TPRs, and it is improper to correlate TIR anomalies with earthquakes without excluding these influences. In this paper, relevant studies will be tested according to the prior probability, random earthquake catalogs and Molchan diagram weighted by the Relative Intensity (RI) index. Our results show that the high TPR in the previous studies were caused by overly large determining windows, and the result was close to random guessing, which means that the extracted TIR anomalies make no contributions to predicting the timing, epicenter and magnitude of earthquakes. By comparing the original Molchan diagram and the Molchan diagram weighted by the RI index, we found that simply taking the fraction of space-time occupied by alarm may give a wrong evaluation for alarm. Moreover, long-term statistical evidence proving the correspondence between earthquakes and TIR anomalies in Greece and Taiwan are still absent. [ABSTRACT FROM AUTHOR]