Research Unit 4 "Natural Hazard" (RU4)

Toward region and site-specific ground-shaking models

The evaluation of potential ground-shaking is a major component of seismic hazard evaluation. The exponential growth of available seismological data was used to improve regional and site specific ground-shaking models (GSM). We developed physics-based and data-driven GSM with a focus on urban areas and stable continental regions (Bindi et al. 2014, Kotha et al. 2016, Drouet & Cotton 2015), both for building normative purposes and for site-specific studies. Developments include high frequency attenuation, ground-motion variability and epistemic uncertainty analysis. A specific focus in our site specific studies was the local shaking amplification of the surface layers (Ktenidou et al. 2015), the surface topography and the soil-structure interaction including the back-scattered energy from buildings. We developed new methods to forecast the site amplification effects in real time (Pilz & Parolai 2016).

Probabilistic hazard and risk models, building codes

In the coming decades, extreme events which previously had little impact will be affecting urban agglomerations of several millions of people. There is an urgent need to develop up-to-date, authoritative and robust building codes. Thus, we interact with the engineering community and contribute to seismic building code committees (DIN, EC8). We have developed the new Probabilistic Hazard Model for Germany to be integrated in the future German seismic building code. We also developed the new seismic hazard and risk model for the Kyrgyz Republic in cooperation with industry and the World Bank and we contributed to the development of the latest European seismic hazard model. The development of reliable hazard and risk models is only possible if key information such as event catalogues, ground-motion data and site conditions are available and continuously updated. The GFZ is strongly engaged in this effort; during the last years we published homogenized earthquake catalogues for Europe (Grünthal et al. 2013) and Central Asia.

Anthropogenic earthquakes

To mitigate the induced seismic hazard we used thermo-hydro-mechanical (THM) models and linked these with statistical methods to the Forward Induced Seismic Hazard Assessment scheme (FISHA) (Hakimhashemi et al. 2014). In collaboration with RU5 "Georessources" we could confirm the predictions of the THM models and FISHA with a unique experiment in the underground rock laboratory Äspö in Sweden, where we performed multiscale injections to monitor the growth and emplacement of hydrofractures in great detail (Zang et al. 2017). The cyclic stimulation scheme of loading and unloading the fracturing net pressure led to a lower accompanied seismicity but simultaneously increased the permeability of the treated rock intervals. From all these experiences a specially adjusted protocol for a field experiment to be carried out at the geothermal site in Pohang, Republic of Korea, was developed within the European funded project DESTRESS.

What are the limits of predictability and how can seismic hazard models be rigorously tested?

We aim to understand the limits of predictability of seismic hazard models and how hazard models can be rigorously tested. Over the last decade, the Collaboratory for the Study of Earthquake Predictability (CSEP) made investigating the forecasting power of seismicity models an integral part in earthquake hazard assessment. As a key participant in CSEP, we continuously contribute to such tests. Within the framework of the Global Earthquake Model (GEM), we have expanded the prospective testing procedures of CSEP to all components of seismic hazard assessments: intensity prediction equation, ground-motion prediction equations, and hazard models (Mak & Schorlemmer 2016). GFZ is taking a major role in GEM and CSEP to provide a key testing center for all types of hazard-relevant models and for investigating the validity of basic concepts employed in seismic hazard research.

Developments in monitoring, fast analysis and rapid response

Impact forecasting

Understanding risk entails firstly understanding the damaging mechanisms of natural perils, and being able to predict their adverse consequences in quantitative and reliable frameworks.

Our group in particular deals with:

  • modeling the potential impact of earthquakes (and other natural perils) in terms of damage to physical assets, loss of life and livelihoods, economical consequences and functional disruption of infrastructure;
  • understanding the underlying uncertainties in the modeling process, and the role of these uncertainties in the subsequent risk management phase;
  • devise innovative methodologies to efficiently collect and integrate the information needed in order to reliably carry out risk assessment and impact forecasting at different spatial scales.

Early Warning and Rapid Response

Early warning refers to the early detection of an event whose unfolding may result in damage and loss, and the subsequent issuing of an alarm that can be used by civil protection authorities to undertake emergency prevention or mitigation actions. In the case of earthquakes, this usually involves installing a dense network of monitoring stations, an endeavour associated with significant economic and technical investment.

Our group is active in the research and development of innovative solutions for Earthquake Early Warning (EEW) that may be scaled according to the available resources, and which also finds applications in economically developing countries. Furthermore, we are advocating for the more tightly knitted integration of rapid impact forecasting into early warning systems in order to support and complement rapid response activities.

Outreach and transfer of knowledge