Historical Seismograms - Evidence from the AD 2000 Izu Islands ...

238 G. A. Eibynorth-west; but subsidences in this valley that were asserted to have been the resultof this shock, and had been ‘explained” in two papers by an internationallycelebrated geomorphologist are clearly shown by well-dated watercolours to haveexisted beforehand. Mistaken traditions of this kind, especially when accepted bywell-known technical authorities, have serious consequences for seismic zoning, andprovide a strong practical argument for rigorously conducted historical studies.Similar confusion and mis-interpretation surrounds the 1855 Wairarapa earthquake.With the exception of a half-legendary event in pre-European times, it isNew Zealand’s largest known earthquake, and occupies an important place in seismologicalhistory because of the accounts of it to be found in later editions of SirCharles Lyell’s “Principles of Geology”, one of the canonical scriptures of Victorianscience, and one of the earliest discussions of the relationship between earthquakesand geological faulting.Lyell’s information came from three New Zealanders he met in England. Twoof them had some geological knowledge, and the third was a military engineer whohad already submitted an official report on the event to his superiors. When theshock occurred he was supervising coastal road-works within a few kilometres of theprobable epicentre. He describes the extent and amount of coastal uplift, but makesno mention of any kind of faulting. Of Lyell’s other informants, one had made hisonly visit to the district before the earthquake, and the other had moved from it tothe epicentral region of the 1848 earthquake, of which he left accounts, before 1847.It was recognized by Ongley (1943) that Lye11 had confused the two events, but notthat the man on whose authority an improbably great length of fresh faulting hasbeen generally accepted could not himself have seen it, even supposing it to havetaken place at that time.The study of historic earthquakes cannot be confined to the examination of writtenevidence. It must be accompanied by field-work; but it is important not toassign historic dates to geological occurrences when written evidence is lacking. Itis also important to remember that statements in quite recent documents may notmean what they would mean had they been written today. Usage and contextchange very quickly. When studying the 1855 shock, I found it helpful to followabout 150 kilometres of the trace of the West Wairarapa Fault on foot and to visitall the places mentioned in contemporary accounts. I soon found that many namesappearing on modern maps are not in the places to which they were applied in thepast. It is essential to consult maps of the same period as the historical recordsbeing studied.The two tsunami-generating shocks in 1946 (Eiby, 1982a, 1982b) occurred ata time when the Observatory was without the resources to undertake field work.Every effort was made to collect information from eye-witnesses, but there was noopportunity to make a proper assessment. The absence of a full study was realizedonly in the course of revising the instrumental data. New Zealand tsunamis oflocal origin are very rare. Since the associated earthquakes were not large, it wasassumed that the height of the wave had been greatly exaggerated, but it proved tobe established by photographs of damage to structures on beach and river terraces,and reports of seaweed caught in electric power-lines. The seismograms showed theearthquakes to have been unusual, and the wave is now attributed to the explosiveexpulsion of mud from a ruptured diapiric fold, a phenomenon that has severaltimes been obsened at inland places nearby.