What do we really know about Historical Earthquakes in Australia?

Given the projected population and economic growth in urban and regional Australia over the coming decades, a robust knowledge of past seismicity must be required for modern seismic hazard assessments. This is important because it influences the recurrence rates used in seismic source-rate models needed to characterize ground-motion hazard. Together with observations from a short modern instrumental record and a growing paleoseismic dataset, information on past seismicity in Australia is gleaned from records of historical earthquakes since the start of European colonization in the late 18th century. As with any scientific data, such documentary evidence extracted from newspapers and other written materials is subject to uncertainties and is prone to the repetition of misinterpreted incomplete data stemming from the absence of a modern re-examination of primary sources. We address these issues by consulting primary documentary sources at Australian libraries and archives to construct a dataset of over 4,000 uniformly assessed macroseismic observations (felt shaking effects) for earthquakes since 1788. These uniformly assessed data have led to the identification of discrepancies in the historical catalogue such as in the locations of the 1897 Beachport, 1918 Queensland and the 1954 Adelaide earthquakes for which catalogued epicentral locations are inconsistent with their macroseismic intensity distributions, felt aftershocks, environmental effects, and reviewed instrumental observations. Our study highlights the benefit of the re-evaluation of underutilized pre-instrumental documentary source materials to homogenize both archival and seismological observations, with modern observations, and to quantify uncertainties that are geological, numeric or societal. This will in turn improve our understanding of historical earthquakes in Australia and will benefit future assessments of modern seismic hazard in the country. We note further that the need to document, interpret, and analyse such underutilised pre-instrumental documentary source materials is not restricted to the analysis of historical earthquakes alone, but is also of benefit to the evaluation of other natural hazards not only in Australia but also around the world.