Comparison of decay features of the 2006 and 2007 Kuril Island earthquake tsunamis

Characteristics of the tsunami decay processes are examined following the definition and discussion of the major terms involved: moving root mean squared amplitude, tsunami coda and non-dimensional tsunami amplitude (NDA). Tsunami waveform data observed at tidal stations in Japan during two Kuril Island earthquake tsunamis in 2006 and 2007 are analysed. The tsunami codas of these events showed the following characteristics. (1) The decay process determined by the tsunami coda of the 2006 tsunami implies that the decay time constant at a tidal station depends on which ocean the station is facing. (2) Fluctuations of the NDAs in the tsunami codas are stochastically approximated by either the Rayleigh distribution or the normal distribution and the distributions of the NDAs at most tidal stations are similar. (3) Differences in the decay processes between these two tsunami events were small with regard to the averaged decay time constants and the distributions of the NDAs. These findings may be the first step towards improved forecasting of the tsunami decay process.