Barberopoulou, A.; Ristau, J.; D'Anastasio, E.; Wang, X. 2013 Sources of information for tsunami forecasting in New Zealand. Lower Hutt, N.Z.: GNS Science. GNS Science report 2013/32 64 p.
Abstract: Tsunami science has evolved considerably in the last two decades due to technological advancements which also helped push for better numerical modelling of the tsunami phases (generation to inundation). The deployment of DART® buoys has also been a considerable milestone in tsunami forecasting. Tsunami forecasting is one of the parts that tsunami modelling feeds into and is related to response, preparedness and planning. Usually tsunami forecasting refers to short-term forecasting that takes place in real-time after a tsunami has or appears to have been generated. In this report we refer to all types of forecasting (short-term or long-term) related to work in advance of a tsunami impacting a coastline that would help in response, planning or preparedness. This report looks at the standard types of data (seismic, GPS, water level) that are available in New Zealand for tsunami forecasting, how they are currently being used, other ways to use these data and provides recommendations for better utilisation. The main findings are: • Current investigation of the use of seismic parameters quickly obtained after an earthquake, have potential to provide critical information about the tsunamigenic potential of earthquakes. Further analysis of this kind of method should be undertaken to determine a path to full implementation. • Network communication of the GPS array is not currently at a stage that can provide data early enough for tsunami warning. It is believed that it has potential but changes including additional staffing may have to happen before major changes are made to the data that is currently provided. • Tide gauge data is currently under-utilised for tsunami forecasting. Spectral analysis (briefly presented in this report), modal analysis based on identified modes and arrival times extracted from the records can be useful for forecasting. • The current study is by no means exhaustive of the ways the different types of data can be used. This publication is only presenting an overview of what can be done. More extensive studies with each one of the types of data collected by GeoNet and other networks should be investigated either through research or through follow-ups to this report. (auth)