Domokos Sármány

Team Leader Model Data Services
Forecast, Development, Model Data Services

Summary:

Domokos Sármány has more than 15 years of academic and industrial expertise in scientific computing, numerical modelling and message-driven data routing. Domokos is currently responsible for leading the development and integration of MultIO -- a message-driven data-routing framework of data output and on-the-fly post-processing for Earth-systems models, in particular ECMWF's weather forecasting system.

Professional interests:
  • Scientific computing
  • High-performance computing
  • Numerical methods
  • Message-driven data workflows
  • Asynchronous I/O
  • Systems of mixed partial differential and algebraic equations
  • C++ code modernisation (C++11/14/17/20)
Career background:

Domokos received his MSc degree in meteorology from Eötvös Loránd University, Budapest, Hungary. His thesis work was on numerical weather prediction, in which he investigated the computation treatment of boundary conditions for limited-area models. He went on to obtain a PhD degree in numerical analysis from the University of Twente, the Netherlands. As part of his research there, he developed and analysed discontinuous Galerkin finite-element methods (DG-FEM) for the effective simulation of electromagnetic wave propagation.

Domokos continued his career as a post-doctoral researcher at the University of Leeds. His area of interest was to create and analyse a novel numerical finite-element-type modelling technique -- called residual distribution -- for the simulation of fluid dynamics, in particular for applications that involve shock capturing and/or atmospheric effects.

Domokos moved to the private sector in 2013 and joined Process Systems Enterprise, now part of Siemens. He was a member of the team that oversaw the numerical engine that underlay the company's flagship gPROMS platform. The engine solves mathematical problems that arise from scientific computing: large sparse systems of differential equations, mathematical optimisation, parameter estimation and sensitivity analysis.

Domokos joined the Development Section at ECMWF's Forecast Department in April 2018. His current responsibilities include overseeing the development and delivery of MultIO, a set of software libraries for efficient data output and on-the-fly post-processing. The main focus is on operational weather forecasting, as well as extreme-weather and climate simulations for Destination Earth, a flagship initiative of the European Union (EU). Domokos has also previously led ECMWF's participation in the EU-funded Maestro project, an effort to co-develop and co-design a middleware for data movement in complex storage and memory architectures.

External recognitions

Member of the Institute of Mathematics and its Applications

Peer-reviewed journal articles

  • C. Haine, U. Haus, M. Martinasso, D. Pleiter, F. Tessier, D. Sármány, S. Smart, T. Quintino, A. Tate. High Performance Computing: ISC High Performance Digital 2021 International Workshops, Frankfurt am Main, Germany, June 24–July 2, 2021, Revised Selected Papers 36, 346-357.
  • D. Sármány and M.E. Hubbard and M. Ricchiuto. A moving mesh implementation of upwind residual distribution. Comput. Math. Appl., 5:1561-1589.
  • D. Sármány and M.E. Hubbard. Upwind residual distribution for shallow-water ocean modelling. Ocean Model., 64:1–11.
  • D. Sármány, M.E. Hubbard and M. Ricchiuto. Unconditionally stable space-time discontinuous residual distribution for shallow-water flows. J. Comput. Phys., 253:86–113.
  • D. Sármány, M.A. Botchev, and J.J.W. van der Vegt. Time-integration methods for finite element discretisations of the second-order Maxwell equation. Comput. Math. Appl., 3:528–543.
  • D. Sármány, F. Izsák, and J.J.W. van der Vegt. Optimal penalty parameters for symmetric discontinuous Galerkin discretisations of the time-harmonic Maxwell equations. J. Sci. Comput., 44(3):219–254.
  • D. Sármány, M.A. Botchev, and J.J.W. van der Vegt. Dispersion and dissipation error in high-order Runge-Kutta discontinuous Galerkin discretisations of the Maxwell equations. J. Sci. Comput., 33(1):47–74.

Technical reports

  • D. Sármány, M.A. Botchev, J.J.W. van der Vegt and J.G. Verwer. Comparing DG and Nédélec finite element discretisations of the second-order time-domain Maxwell equation. Technical Report 1912, Department of Applied Mathematics, University of Twente, Enschede, December 2009.
  • D. Sármány, F. Izsák, and J.J.W. van der Vegt. High-order accurate discontinuous Galerkin method for the indefinite time-harmonic Maxwell equations. Technical Report 1889, Department of Applied Mathematics, University of Twente, Enschede, January 2009.