Lam, Dan-Thuy

PhD student


In a context of increasing model-based decision-making, reliability of model predictions is even more critical. It turns out that the calibration process upon which predictions of interest are obtained is still not as efficient in practice in spite of being essential in the modelling workflow. With the currently observed tendency towards greater conceptual model complexity, large parameterization of distributed models indeed often results in long simulation runtimes that consequently make model calibration particularly challenging. In the light of recent developments of inverse algorithms in hydrogeology, the main purpose of this thesis is to focus firstly on several selected inverse approaches for the calibration and predictive uncertainty assessment of a large-scale numerical groundwater flow model in transient state. Then, techniques to achieve computational efficiency, without sacrificing accuracy, of these normally time-consuming procedures will be investigated. On the basis of a real-world inspired synthetic case, we will try to develop an efficient calibration strategy allowing uncertainty quantification before moving on to the real problem.



  • 2011 – 2014: Engineer’s degree (eq. MSc) from the School of Engineering Geophysics of the EOST (School and Observatory for Earth Sciences), University of Strasbourg. Course in geophysics and hydrogeology. Thesis: “Numerical modelling of flow and transport in unsaturated and saturated porous media using MARTHE software”
  • 2009 – 2012: Licence (eq. BSc) in Earth, Space and Environmental Sciences, EOST, University of Strasbourg




Office: E320

Phone: +41 (0) 32 718 26 92


Centre for Hydrogeology and Geothermics (CHYN)
Emile-Argand 11
CH-2000 Neuchâtel