Abstract. Space-time TIR anomalies, observed from months to weeks before earthquake occurrence, have been suggested by several authors as pre-seismic signals. Up to now, such a claimed connection of TIR emission with seismic activity has been considered with some caution by scientific community mainly for the insufficiency of the validation data-sets and the scarce importance attached by those authors to other causes (e.g. meteorological) that, rather than seismic activity, could be responsible for the observed TIR signal fluctuations. A robust satellite data analysis technique (RAT) has been recently proposed which, thanks to a well-founded definition of TIR anomaly, seems to be able to identify anomalous space-time TIR signal transients even in very variable observational (satellite view angle, land topography and coverage, etc.) and natural (e.g. meteorological) conditions.

Its possible application to satellite TIR surveys in seismically active regions has been already tested in the case of several earthquakes (Irpinia: 23 November 1980, Athens: 7 September 1999, Izmit: 17 August 1999) of magnitude higher than 5.5 by using a validation/confutation approach, devoted to verify the presence/absence of anomalous space-time TIR transients in the presence/absence of seismic activity. In these cases, a magnitude threshold (generally M<5) was arbitrarily chosen in order to identify seismically unperturbed periods for confutation purposes.

In this work, 9 medium-low magnitude (4<M_b<5.5) earthquakes which occurred in Greece and Turkey have been analyzed in order to verify if, even in these cases, anomalous TIR transients can be observed.

The analysis, which was performed using 8 years of Meteosat TIR observations, demonstrated that anomalous TIR transients can be observed even in the presence of medium-low magnitude earthquakes (4<M_b<5.5). As far as the research (just started) of possible correlation among TIR anomalies and earthquake occurrence is concerned, such a result suggests that: a) in order to identify seismically unperturbed periods for confutation purposes, a magnitude threshold (at least) lower than 4 should be used; b) the proposed validation/confutation approach should be applied in low-seismicity areas in order to find suitably long seismically quiescent periods.