Quantitative lacustrine paleoseismology in Carinthia

Despite being considered a country of moderate seismicity, several historical earthquakes in Austria caused massive impact on infrastructures and landscape. Seismological data and historical records generally do not extend far enough back in time to make correct estimates of where and how often such catastrophic events could take place. It is possible that very strong earthquakes took place in prehistorical times and are thus overlooked in the present estimates of seismic hazard. As a consequence, future events may come with an unexpected large impact to (critical) infrastructure and excessive economic and human losses. We overcome this knowledge barrier by using lake sediments as natural seismometers. Strong earthquakes can deform lacustrine sediments and produce underwater landslides. Due to the continuous accumulation of mud on the lake bottom, such evidence of past earthquakes gets preserved and can get retrieved by geophysical techniques and sampled by sediment cores. The Quake-Lake Carinthia project applies this paleoseismic research strategy on the large lakes in Carinthia, one of Austrias most seismic areas. Due to its unique setting -where many lakes have been impacted by several well-documented historical earthquakes- we determine relationships between the strength of shaking and the characteristics of the sedimentary features. In this way, our records do not only show when strong earthquakes took place in the last 14.000 yrs, but also show how strong they were and in which area their epicenter was located. These long earthquake archives will also generate new insights about how earthquake activity changed over time as a consequence of the melting of the alpine icecap since the last Ice Age. Paleoseismological research is relatively new in Austria, and lake sediments as a natural earthquake archive have not been studied there. We implement and improve the newest methods in lacustrine paleoseismology: i) in-situ geotechnical tests to calculate how strong shaking needs to be to destabilize underwater slopes, ii) different geochemical and physical proxies to better understand the nature of the earthquake impact on lake sediments and surrounding landscapes. The detailed work on the Carinthian lakes forms the fundaments of a worldwide comparison of lake paleoseismology, where we evaluate how different sedimentary features in different lake systems can be used to reveal the shaking strength of past earthquakes. We combine present and new knowledge to construct a generic methodology which can be applied to improve seismic hazard assessments in many places worldwide.

Seismological data generally do not extend far enough back in time to make correct estimates of where and how often catastrophic earthquakes could take place. As a consequence, future events may cause an "unexpected" large impact and excessive losses. This project used lake sediments in Austria as natural seismometers to reconstruct the occurrence, location and strength of earthquakes over the past 14000 years. For Carinthia, we found that the 1348 AD earthquake was an exceptionally strong event as it generated the strongest shaking of the last 10000 years in that region. It is only superseded by a larger event that took place ca. 13500 years ago and which completely altered the bottom of Woerthersee. Moreover, we found the first evidence for seismic bursts, i.e. periods of a few 100 years long in which many strong earthquakes occur in short succession. This finding implies that seismic hazard may be elevated in the years and decades after a strong earthquake takes place. Moreover, we found relationships between how strong the earthquake shaking is, and the size and type of evidence in the lake sediments. Based on these findings, we reconstructed the occurrence and shaking strength for earthquakes in the past 2,800 years, which revealed that the national seismic hazard map is reliably representing the seismic hazard in central Carinthia. For Tyrol, the lake sediment data let us infer that several large prehistoric rockslides were triggered by strong earthquake shaking. This concluded a long-standing debate on what have caused these dramatic landscape changes. Moreover, we found evidence for an active fault system under Achensee which lastly ruptured around 8300 years ago, producing an earthquake of magnitude ~6-6.3. Our reconstructions show spatial migrations of paleo-seismicity, indicating that seismic hazard is not constant over millennial time scales. Simulations of shaking strength indicated that, in addition to Central Tyrol, also southern Bavaria may experience significant damage during earthquakes that originate near the Inn Valley. Another project pillar are process-oriented studies for a better understanding the different parameters that influence earthquake-induced imprints in lake sediments. By studying lakes in Chile, Alaska, Israel, Taiwan and Austria, we gained insights in the development of sediment disturbances, underwater landslides and mud avalanches. This is crucial to evaluate the reliability and uncertainties in our paleoseismic reconstructions and to discover new strategies to retrieve more information on past earthquakes. We detected that earthquake recurrence patterns can be fundamentally different: i.e. some tectonic plate boundaries exhibit a remarkably regular cycle of strong earthquakes, whereas locations in the interior of plates rather exhibit a random or clustered earthquake recurrence. Such reconstructions advance our understanding of how large earthquakes work and improve our assessment of earthquake hazards in different regions worldwide.