Simulation is routinely used today in many scientific applications in Earth Sciences. For example, it replaces expensive laboratory testing, enables geo-engineers to explore natural resources and allows climate scientists to assess future global and regional climate developments. These examples fit the classical usage of computer simulation: given an abstract mathematical model of a system, one develops a discrete approximate analog that can be processed by a computer and allows studying the system’s behavior.

In the Earth Sciences, however, there is a multitude of situations in which such agreed-upon abstract mathematical models are not available. This may be due to either the absence of sufficiently detailed information regarding the systems’ characteristics, or to the fact that the systems have not, or merely partially, been described as abstract mathematical models thus far.

Improving constraints by (instrumental) observation is in many cases not possible – either because of inaccessibility of remote locations or because the instrumental record does and will not cover a long enough time span. Therefore, in contrast to the classical simulation strategies described above, we suggest simulation to take the role of an explorative tool dedicated to advancing our qualitative and quantitative understanding of Earth’s complex systems in those cases where we lack complete understanding.


Through the exploration of virtual worlds we aim to improve the understanding of ours


This approach will be most relevant in Earth subsystems that are governed by multiple feedbacks, by a large number of components linked by various physical mechanisms with different spatial and temporal characteristics or self organization, in short by systems exhibiting various modes of complexity. Among these examples we have chosen fields that have particular relevance for humankind on a dynamic planet. These are solid earth dynamics and earthquakes, atmosphere-hydrosphere dynamics and climate evolution, and hydrological flow and transport processes. We foresee studying these fields with a variety of mathematical and physical modeling strategies hoping to contribute to the understanding of the intricate complexity of the systems’ functioning.

Three interwoven Research Areas will form the focus of the research school. They are in turn applied to the three topical research fields in Earth Sciences mentioned above.


In the collaboration in GEOSIM, the partners at GFZ, Potsdam University and FU Berlin will systematically link methodological expertise from the Earth and Mathematical Sciences in all individual studies.  The PhD fellows play an active role by co-defining their own research as well as their involvement in the organization of seminar series and workshops.

The individual projects to be established within this program will all be led by more than one principal investigator from at least two of the participating scientific disciplines. Each Ph.D. student will thus be co-advised by two professors from different disciplines.

The curriculum of GEOSIM will be developed in collaboration with partners and fellows and is dedicated to training a new generation of young, interdisciplinary scientists with skills in combining applied geo-science with the development of physical concepts of Earth systems, mathematical abstraction, and the use of high-performance computing.