Jonas Sukys

Dr. Jonas Sukys

Head of the Scientific Computing group

Department Systems Analysis, Integrated Assessment and Modelling

About Me

I am head of the Scientific Computing group,
at the Department of Systems Analysis, Integrated Assessment and Modelling.

Research interests

[all publications: Google Scholar]

  • Massively parallel high performance computing (HPC), scalable parallelization, emerging computing platforms
  • Uncertainty Quantification and Propagation (UQ+P) for deterministic and stochastic models
  • Multi-level Monte Carlo (MLMC) methods for optimal hierarchical variance reduction in UQ+P
  • Hyperbolic nonlinear partial differential equations (PDEs): shallow water, multi-phase cavitation dynamics

Open positions

In early 2018 two-year postdoctoral positions are expected to be available in SPUX and DATALAKES projects - please contact me if you are interested (no applications yet).

Research projects


Parallelization and deployment of scalable Particle Markov Chain Monte Carlo (PMCMC) framework implementing Bayesian uncertainty quantification for stochastic models in environmental data sciences. Multi-level variance reduction techniques and dynamic parallel load balancing are employed to mitigate significant computational requirements. Applications include realistic river invertebrate mesocosms using Individual Based Models (IBMs) and SDEs, ground water flow propagation using SPDEs, and stochastic 2-D urban flood analysis.

Collaborations: Peter Reichert, Nele Schuwirth (SIAM, Eawag) , Mira Kattwinkel (U Koblenz-Landau), Fabrizio Fenicia (SIAM, Eawag), Mario Schirmer (W+T, Eawag), Jörg Rieckermann, Joao Leitao (SWW, Eawag), Anze Zupanic, Piet Spaak (UTOX, ECO, Eawag), Torsten Hoefler (SPCL, ETH Zürich), Siddhartha Mishra (SAM, ETH Zürich).


Bayesian Inference for Geneva and other Lakes with remote sensing and in-situ measurement data assimilation using Kalman and particle filters, empirical dynamics methods and large scale numerical simulations of hydrological and ecological models. Multi-level variance reduction techniques are employed to allow for realistic hydrological model resolutions while keeping accurate uncertainty estimates.

Collaboration: Damien Bouffard (SURF, Eawag), Johny Wuest (EPF Lausanne), Siddhartha Mishra (SAM, ETH Zürich) and SDSC.


Parallelization potential study for the Hamiltonian Monte Carlo methods with potential applications in water catchment modeling and solar physics.

Collaboration: Carlo Albert (SIAM, Eawag), Simone Ulzega (ZHAW), Antonietta Mira (USI Lugano) and SDSC.


Coupling of the PyMLMC parallel sampling framework and the non-hydrostatic multi-layer solver for landslide induced tsunami wave propagation. Movie: link.

Collaboration: Manuel Castro (CADMOS, U Malaga, Spain) and Siddhartha Mishra (SAM, ETH Zürich).


Uncertainty quantification using optimal fidelity multi-level Monte Carlo for large scale direct numerical simulations of cloud cavitation collapse. Partly supported by CSCS, PRACE and INCITE allocations of supercomputer access to PIZ DAINT, JUQUEEN, FERMI and MIRA. Movies: SC15, SC16.

Collaboration: Ursula Rasthofer, Fabian Wermelinger, Panagiotis Hadjidoukas and Petros Koumoutsakos (CSElab, ETH Zürich).

{{ element.title }}


How to investigate the spatial variability in lakes?

{{ element.title }}

{{ element.title }}

{{ element.title }}

{{ element.title }}


Phone: +41 58 765 5310
Fax: +41 58 765 5802
Address: Eawag
Überlandstrasse 133
8600 Dübendorf
Office: FC D10

{{ element.title }}

{{ element.title }}

{{ element.title }}

{{ element.title }}

{{ element.title }}