# Current students

On this page, the current GeoSim students present their projects.

## RA I, Solid Earth

### Felix Hoffmann (formerly Eckelmann)

**Education**

MSc in GeoScience (2013) at the University of Potsdam. Thesis title: 'Recent crustal deformation in the orogenic pateau interior: InSAR observations from the Salar de Pocitos, NW Argentina'.

**Current project with GeoSim (since 01.10.13) **

The preliminary title of my PhD thesis is ‘Deformation patterns related to the seismic cycle in the Central Andes using geodetic data.’ I am working at the GFZ Potsdam, section 3.1 Lithosphere Dynamics.

Supervisor: Prof. Dr. Onno Oncken (GFZ)

The western coast of the South American continent is dominated by the subduction of the Nazca plate. This special tectonic regime causes a high seismic and also volcanic activity with huge megathrust earthquake events of magnitudes of Mw=8 and even larger. In order to improve the knowledge about the tectonic mechanisms behind these events my goal is to apply different geodetic methods as GPS and InSAR in the northern part of Chile and northern Argentina. I will investigate regions at different stages of the seismic cycle and analyze especially the inter- and post-seismic behaviour of the crustal deformation.

### Stefan Rüdrich

**Education**

MSc in Mathematics (2012) at Freie Universität Berlin. Majors: Discrete Mathematics, Stochastic Dynamics. Thesis title: 'Identification of clusters in graphs - non-spectral methods, the random walk approach and their generalization to directed graphs'

**Current project with GeoSim (since 15.01.2013)**

'Metastability in Time Series Analysis' at the Department of Mathematics and Computer Science, Freie Universität Berlin.

Supervisors: Christof Schütte (FU Mathematics) and Sebastian Hainzl (GFZ 2.1).

To analyze and forecast geoscientific processes, methods of network analyses are widely used, exploiting the relation between metastable substructures of a recurrence network and the topology of the modeled system. Usually it is assumed that the future behavior of these processes does not depend on past behavior, but only on their current state. That might not be justified for many cases in question, e.g., there is no evidence that the occurrence of earthquakes was unrelated to former events and observation rather suggests the opposite.

In this context, processes with memory and memory-coupled networks were not yet thoroughly explored. Abandoning the above assumption, Time Series Analysis tools need to be adapted for analyzing non-reversible processes. Consequently, it would be obvious to ask for the interpretation of clusters and other network substructures with regard to the original processes.

**1st author publications**

Rüdrich, S., and Sarich, M. and Schütte, C. (2017): Utilizing hitting times for finding metastable sets in non-reversible Markov chains. Journal of Comp. Dynamics, accepted for publication

## RA II, Solid Earth

## RA II, Atmosphere-Hydrosphere

### Annette Müller

**Education**

Diplom in Mathematics (2011) at TU Berlin. Thesis title: 'Das konforme Modul eines Knotens am Beispiel des symmetrischen Kleeblattknoten'.

**Current project with GeoSim (since 01.04.2012)**

'A transition scenario from three to two-dimensional regimes of turbulence based on the energy-vorticity theory of fluid mechanics using an infinite cascade of helicity related constraints' at the Institute of Meteorology, FU Berlin.

Supervisors: Peter Nevir (FU Meteorology)and Sebastian Reich (UP Mathematics).

Especially from the meteorological point of view, the challenge is to combine the different 2D-and 3D-turbulence models. A tornado is an example of a meteorological event appearing in the 3D convective scale. Those events are characteristic for three-dimensional turbulence and can be modelled as vortex filaments. Two-dimensional turbulence theory contains larger-scale processes. For example low-pressure and high-pressure areas. These synoptic processes can be described by point vortices. The energy-vorticity theory will be applied to combine these different models.

**1st author publications**

Müller, A., and P. Névir. "A geometric application of Nambu mechanics: the motion of three point vortices in the plane." Journal of Physics A: Mathematical and Theoretical 47.10 (2014): 105201.

Müller, Annette et al. Applications of point vortex equilibria: blocking events and the stability of the polar vortex. Tellus A (2015), 67, 29184, http://dx.doi.org/10.3402/tellusa.v67.29184

**Awards**

Outstanding Student Poster Award, European Geosciences Union (EGU) General Assembly, 2013

### Isabel Urrutia

**Education**

Geology specialization (2014) at University of Concepcion, Chile. Thesis title: ‘Mega-thrust seismic cycle at Isla Santa María from numerical modelling of crustal deformation before, during and after the Mw8.8 Maule 2010 earthquake’.

**Current project with GeoSim (since 01.04.2014)**

‘Linking deformation patterns at various spatiotemporal scales along the South-American subduction zone’.

Supervisors: Onno Oncken (GFZ 3.1) and Ralf Kornhuber (FU Mathematics).

I will compile published GPS displacements and use time-series of the IPOC database to characterize the inter-seismic deformation and construct a realistic 3D FE-Model of the South American subduction zone incorporating all the available geometrical and rheological information to estimate the locking degree and its variation along the margin using a single consistent model allowing direct comparison of result.

I will analyze deformation patterns of the force-arc, manifested in the geology and geomorphology in a millennial time scale. I must choose a segment(s) along South American subduction margin that is best suited to this study. Suitable segments should have sufficient geo-chronological data and geological features for geomorphological analysis, e.g. drainage basins, marine terraces, peninsulas, etc. I will compile existing gradients in uplift rates complemented with detailed study of morphometric indices of drainage basins to postulate gradients of uplift rates and to distinguish different seismotectonic segments.

By comparing results of two analysis I will estimate the main parameters controlling the expression of the short-term inter-seismic deformation in the differential terrace uplift and margin segmentation.

## RA II, Hydrological Flow

### Weishi Wang

**Education**

Msc in Water Resource Engineering （2011） at Katholieke Universiteit Leuven, Belgium.Thesis title: 'The detention area and the concave green space study in the city of Turnhout, Belgium'.

**Current project with GeoSim (since 01.04.2014)**

'Surface water and shallow groundwater exchange modelling'.

Supervisors: Sascha Oswald (UP Hydrology)

My research is try to use some method to increase the modelling accuracy and have a better understanding about the surface water and shallow groundwater interaction.

### Ulrich Wilbrandt

**Education**

MSc in mathemathics / scientific computing (2011) at FU Berlin. Thesis title: 'A posteriori Optimierung von Parametern in stabilisierten Finite-Elemente-Methoden für Konvektions-Diffusions-Gleichungen'.

**Current project with GeoSim (since 01.10.2011)**

We study coupled Stokes and Darcy equations which for example model surface water / ground water interactions. We will particularly focus on implementation, analysis, inf-sup stable versus stabilized discretizations. Furthermore we will apply these methods to real world problems from geophysics. This includes simulations in three space dimensions and time dependencies. We hope to compare our results to existing commercial codes available.

Supervisors: Volker John (WIAS) and Sascha Oswald (UP Meteorology).

**1st author publications**

Wilbrandt, U., Bartsch, C., Ahmed, N., Alia, N., Anker, F., Blank, L., ... & Meesala, R. (2017). ParMooN-A modernized program package based on mapped finite elements. Computers & Mathematics with Applications.

## RA III, Solid Earth

### Annabel Händel

**Education**

Diploma in Geophysics (2012) at the Institute of Earth- and Environmental Sciences of the University of Potsdam. Thesis title:'Data driven selection and ranking of ground-motion prediction equations for subduction zones'.

**Current project with GeoSim (since 01.04.2012)**

'Determination of attenuation properties from ambient vibrations' at the Institute of Earth- and Environmental Sciences, University of Potsdam.

Supervisors: Matthias Ohrnberger (UP Geophysics & Seismology), Frank Scherbaum (UP Geophysics & Seismology), Volker John (WIAS).

Understanding how earthquake waves are attenuated in the ground is crucial for any seismic hazard analysis. Especially the high-frequency filtering effect of the uppermost layers (described by the parameter kappa) can have a considerably influence on the proper prediction of future ground motions and needs to be assessed.

The main goal of this reserch project is the evolution of a non-destructive method to determine attenuation from noise data. We will invesitgate how seismic interferometry can be used to extract amplitude and attenuation information from ambient vibrations. We will further simulate a homogeneous scattering medium using finite differences to validate the obtained results and to better understand the mechanisms of seismic wave attenuation in the ground.

**1st author publications**

Haendel, A., Specht, S., Kuehn, N. M., and Scherbaum, F. (2015). Mixtures of ground-motion prediction equations as backbone models for a logic tree: an application to the subduction zone in Northern Chile. Bulletin of Earthquake Engineering, 13(2), 483-501.

Haendel, A., Ohrnberger, M., and Krüger, F. (2016). Extracting near-surface QL between 1-4 Hz from higher-order noise correlations in the Euroseistest area, Greece. Geophysical Journal International, 207(2), 655-666.

### Stefan Mauerberger

**Education**

Diploma in physics (2011) at Ludwig-Maximilians-Universität, Munich with a minor in Electronics and Geophysics. Thesis title: 'Investigation of Deep Earth Wave Propagation with the Spectral-Element Method in Spherical'.

**Current project with GeoSim (since 01.10.13)**

'Combining homogenization and Bayesian statistics by incorporating apriori information for downscaling results in full waveform inversion'.

Supervisor: Matthias Holschneider (UP Mathematics) and Serge Shapiro (FU Geophysics).

In my current project I investigate how to incorporate a-prioryinformation into the framework of full waveform inversion.

A-priory information can be taken from other disciplines, experiments or approaches.

As an example each, this might be limits for the elastic parameters from rock physicist, actual data given by a well-log or well known features obtained through high frequency approximations.

To introduce these information into the 'blurred' model obtained through full waveform inversion homogenization theory and Bayesian statistics are promising approaches.

## RA III, Atmosphere-Hydrosphere

### Forough Marzban

**Education**

MSc. in Computer Engineering -Artificial Intelligent- (2009) at Islamic Azad University Science and Research branch, Tehran, Iran. Thesis title: 'Identification and Predication of Chaotic time series by using Dynamic Neural Network Discrete time'.

**Project with GeoSim (will starts on 01.06.2014)**

'Estimation of high resolution 2m Temperature in Urban areas using Artificail intellegence methods: case study Berlin'

Supervisors: Sahar Sodoudi (FU Meteorology) and Lutz Prechelt (FU Computer Science).

It is necessary to develop a method for the estimation of air temperature with reasonable accuracy and spatial and temporal resolution in the urban areas with low temperature gauge density. For this reason, we should develop an appropriate machine learning techniques. Thus, in my research, first of all, different machine learning techniques including Neural networks (Dynamic Neural unit , Multilayer perceptrons (by using Back Propagation algorithm)), Bayesian Network, Fuzzy System, Support vector machine are considered and then, by applying each technique, a predicator model of air temperature will be generated. The comparison between these methods will be done, finally we evaluate the accuracy of each model and choose the best one for the high resolution temperature prediction.

**1st author publications**

Marzban, F., Ayanzadeh, R., and Marzban, P. "Discrete time dynamic neural networks for predicting chaotic time series." Journal of Artificial Intelligence 7.1 (2014): 24.

Marzban, F., Sodoudi, S. & Preusker, R. (2018) The influence of land-cover type on the relationship between NDVI–LST and LST-Tair, International Journal of Remote Sensing, 39:5, 1377-1398

### Edoardo Mazza

**Education**

B.Sc with Honours in 'Geophysics and Meteorology' (2013) at the University of Edinburgh. Thesis title: 'An investigation of the influence of different climate variability modes on the Atlantic tropical cyclone activity using a multiple linear regression model'.

**Current Project with GeoSim (since 1.10.2013)**

'Dynamic and thermodynamic control on medicane activity' at the Institute of Meteorology, FU Berlin.

Supervisors: Uwe Ulbrich (FU meteorology) and Rupert Klein (FU Mathematics).

The Mediterranean Sea, under specific circumstances, is an area prone to the development of intense sub-synoptic scale, axisymmetric cyclones, often referred to as “Medicanes”.

Although they share some visual and physical similarities with tropical cyclones, several aspects of their genesis and evolution remain unknown. My research is aimed at untangling the importance of different dynamical and thermodynamical factors throughout the lifecycle of medicanes using high- resolution numerical model simulations.The research may provide useful information to improve the forecasting of such cyclones and hence reduce the associated risk sin highly-vulnerable areas.

**1st author publications**

Mazza, E., Ulbrich, U., and Klein, R., 2017: The Tropical Transition of the October 1996 Medicane in the Western Mediterranean Sea: A Warm Seclusion Event. Mon. Wea. Rev., 145, 2575–2595, doi.org/10.1175/MWR-D-16-0474.1

### Linda Michalk

**Education**

Diplom in mathematics (2011) at Freie Universität Berlin. Thesis title: 'Numerical methods for the linear advection equation: Plateaus vs. Extrema'.

**Current project with GeoSim: (since 01.10.2011)**

'Advanced finite volume schemes numerics for advection dominated problems' at the Institute of Mathematics, Department of Mathematics and Computer Sciences, Freie Universtät Berlin.

Supervisors: Rupert Klein (FU Mathematics) and Uwe Ulbrich (FU Meteorology).

The goal of my work is to compare existing numerical methods, find improvements or develope new ones, that yield satisfying results for solving geophysical problems - in particular advection dominated problems.

## RA III, Hydrological Flow

### Tobias Pilz

**Education**

MSc in geoecology (2014) at University of Potsdam. Thesis title: 'Influence of climate change and management on water availability and erosion in a meso-scale catchment in South Africa'.

**Current project within GeoSim (since 01.03.2014)**

'Hydrological modelling at the management scale -- improving simulation results by data assimilation'.

Supervisors: Axel Bronstert (UP Hydrology) and Sebastian Reich (UP Mathematics).

Hydrological extremes such as droughts and floods have the potential to cause heavy damages on human goods and even threat their lives. Therefore, forecasting of such extremes with adequate lead-time is important to warn people and reduce damages. However, due to highly non-linear interactions hydrological simulations and forecastings are still accompanied by a high level of uncertainty. Quantifying and reducing that uncertainty is the main focus of my dissertation. To achieve this data assimilation techniques - widely used in meteorology but yet poorly recognised for hydrological applications - shall be applied and tested to make use of available observations covering different scales (including in-situ meaurements and remote-sensed data) and improve forecasting of droughts and floods.

**1st author publications**

Pilz, T., Francke, T., and Bronstert, A.: lumpR 2.0.0: an R package facilitating landscape discretisation for hillslope-based hydrological models, Geosci. Model Dev., 10, 3001-3023, doi.org/10.5194/gmd-10-3001-2017, 2017