Team seminars

17th seminar of Team CaliSto “Analyzing the Applicability of Random Forest-Based Models for the Forecast of Run-of-River Hydropower Generation”

  • Date & Time: Thursday, June 16th, 2022 at 10:00 (CET)
  • Speaker: Valentina Sessa (Mines ParisTech – CMA, Sophia Antipolis, France)
  • Topic:
    Analyzing the Applicability of Random Forest-Based Models for the Forecast of Run-of-River Hydropower Generation
  • Abstract :
    Analyzing the impact of climate variables into the operational planning processes is essential for the robust implementation of a sustainable power system. This work deals with the modeling of the run-of-river hydropower production based on climate variables on the European scale. A better understanding of future run-of-river generation patterns has important implications for power systems with increasing shares of solar and wind power. Run-of-river plants are less intermittent than solar or wind, but also less dispatchable than dams with storage capacity. However, translating time series of climate data (precipitation and air temperature) into time series of run-of-river-based hydropower generation is not an easy task as it is necessary to capture the complex relationship between the availability of water and the generation of electricity. This task is also more complex when performed for a large interconnected area.  In this work, a model is built for several European countries by using machine learning techniques. In particular, we compare the accuracy of models based on the Random Forest algorithm and show that a more accurate model is obtained when a finer spatial resolution of climate data is introduced. We then discuss the practical applicability of a machine learning model for the medium-term forecasts and show that some very context-specific but influential events are hard to capture.

16th seminar of Team CaliSto “Inverse control of visual simulations with deep learning”

  • Date & Time: Thursday, May 19th, 2022 at 10:00 (CET)
  • Speaker: Guillaume Cordonnier (GraphDeco Team, Inria SAM)
  • Topic:
    Inverse control of visual simulations with deep learning
  • Abstract :
    The synthesis of visually plausible digital copies of the physical world is critical for several applications, including entertainment, marketing, communication, planning and training, etc. In particular, the simulation of dynamic, physically-driven natural phenomena yields several challenges: the algorithms should quickly and robustly emulate physical phenomena which are inherently complex and chaotic. Furthermore, users need to manipulate the simulations for art, business, or communication, which add a requirement: the solver should be intuitively and efficiently controllable. In this talk, I will highlight the differences and similarities between visual and physical simulations, and I will illustrate the use of machine learning to generate, accelerate, or control those simulations.

15th seminar of Team CaliSto “Lagrangian modeling of intermittency in turbulence”

  • Date & Time: Thursday, April 7th, 2022 at 10:30 (CET)
  • Speaker: Roxane Letournel (SAFRAN, Chateaufort)
  • Topic:
    Lagrangian modeling of intermittency in turbulence
  • Abstract :
    The characterization of intermittency in turbulence is the subject of numerous experimental, numerical and mathematical modeling studies.
    Stochastic processes allow to model the phenomenon from a fundamental point of view, but also to develop methods to study turbulent flows loaded with particles. Inspired by the multiplicative Gaussian chaos theory, we propose a model for the pseudo-dissipation of kinetic energy based on a sum of carefully chosen Ornstein-Uhlenbeck processes. The developed method allows to find existing processes from the literature, to derive new ones with adequate regularizations and to propose a simple and efficient numerical simulation.

14th seminar of Team CaliSto “Numerical simulations assisted by orthogonal eigenvalue decomposition and optimal transport”

  • Date & Time: Thursday, March 24th, 2022 at 14:30 (CET)
  • Speaker: Michel Bergmann (Team MEMPHIS, Inria, Bordeaux)
  • Topic:
    Numerical simulations assisted by orthogonal eigenvalue decomposition and optimal transport
  • Abstract :
    The numerical simulation of multiphysics problems often involves interfaces subjected to large deformations. In this context, we develop methods like Immersed Boundary which allow to get rid of conformal meshes at the interfaces. We will then show how this type of approach can be coupled with reduced models like POD (Proper Orthogonal Decomposition) and with optimal transport approaches to accelerate numerical simulations and to go towards numerical twins from experimental data.
    The applications presented will focus on problems of aerodynamics, bio-locomotion and bio-inspiration.

13th seminar of Team CaliSto “Homogeneization of Markovian process and application to quantum continuous measurement”

  • Date & Time: Tuesday, March 15th, 2022 at 10:00 (CET)
  • Speaker: Raphael Chetrite (Laboratoire J.A. Dieudonné, Nice)
  • Topic:
    Homogeneization of Markovian process and application to quantum continuous measurement
  • Abstract :
    In many physical systems, we are faced with an interaction between mechanisms that evolve at very different time scales. In this seminar, we will consider Markovian processes with such characteristics, and we will exploit the time scale separation to obtain an effective Markovian dynamics on a reduced state space. In particular, compared to existing studies, we will introduce a more general formalism.

    Finally, as a privileged example, we consider the limit of the strong continuous measure of quantum systems described by a Belavkin diffusive equation.

12th seminar of Team CaliSto “From random trajectories to random fields: Is fluid dynamics intrinsically random?”

  • Date & Time: Wednesday, December 15th, 2021 at 11:00 (CET)
  • Speaker: Simon Thalabard (University Côte d’Azur)
  • Topic:
    From random trajectories to random fields: Is fluid dynamics intrinsically random?
  • Abstract :
    This talk intends to briefly introduce the mechanism of spontaneous stochasticity, a radical version of chaos emerging within turbulent environments. Roughly said, the spontaneous stochasticity describes the finite-time emergence of macroscopic randomness out of thermal fluctuations, due to the presence of some type of small-scale roughness.
    I will specifically discuss two simple frameworks where spontaneous stochasticity can be identified, namely the turbulent diffusion of fluid particles and the Kelvin-Helmholtz shear-layer instability.
    Those two examples suggest that fluids at high-Reynolds number are intrinsically random objects, be them observed from the Lagrangian or from the Eulerian perspective.

11th seminar of Team CaliSto “Organic matter in the atmosphere: from micro and nanoplastics to cloud chemistry”

  • Date & Time: Thursday, November 18th, 2021 at 10:00 (CET)
  • Speaker: Angelica Bianco (LaMP/CNRS, Université Clermont Auvergne)
  • Topic:
    Organic matter in the atmosphere: from micro and nanoplastics to cloud chemistry
  • Abstract :
    This seminar will firstly focus on the presence of micro and nanoplastics in the atmosphere, with a brief introduction of the experimental determination of the concentration and the technological limitations. The second part will be focused on reactivity, especially concerning the aqueous reactivity with hydroxyl radicals. in the last part we will explore more the composition and transformations happening in atmospheric droplets and their impact on pollution and climate.

10th seminar of Team CaliSto “Fragmentation of brittle fibres in turbulence and CO2 convective dissolution in a 3-D granular porous medium”

  • Date & Time: Tuesday, November 10th, 2021 at 10:00 (CET)
  • Speaker: Christophe Brouzet (CNRS InPhyNi)
  • Topic:
    Fragmentation of brittle fibres in turbulence and CO2 convective dissolution in a 3-D granular porous medium: two experimental studies
  • Abstract:
    In this seminar, I will talk about two experimental works that I have performed recently.
    In the first part, I will discuss the fragmentation of brittle fibres in a turbulent flow, motivated by fundamental questions related to the fate of plastic debris in oceans. Using laboratory experiments complemented by numerical simulations and theoretical analyses, our results demonstrate that the fragmentation process is limited at small scales by a physical cut-off length. This length originates from the fluid-structure interactions between the debris and the surrounding turbulent flow field, and is shown to be independent of the brittleness of the fibres. Such limitation mechanism of the fragmentation process at small scales is able to reproduce the size distribution of floating plastic debris measured in the ocean.
    In the second part, I will present recent results on CO2 convective dissolution in a 3-D granular porous medium, related to the geological storage of CO2 in deep saline aquifers, a promising technique to mitigate global warming. When CO2 is injected in the geological formation, it dissolves partially in the interstitial brine, thus rendering it denser than the CO2-devoid brine below, which creates a convective instability. The instability has been intensively discussed by numerical and theoretical approaches at the Darcy scale, but few experimental studies have characterized it quantitatively. Our experimental measurements of the onset characteristics of this instability in a 3-D porous medium highlight that the growth rate of the instability is in clear discrepancy with the theoretical predictions, while the front of the instability exhibits a multi-scale nature. These results are consistent with a forcing of convection by porosity fluctuations, recently described in a theoretical model.
  • Link to the presentation: TBA

9th seminar of Team CaliSto “Special Gaussian processes & surrogate models”

  • Topic: Two presentations on:
    1) Analysis of complex simulation experiments with Gaussian processes2) Beyond Surrogate Modeling: Learning the Local Volatility via Shape Constraints
  • Date & Time: Friday, July 2nd, 2021 at 10:00 (CET)
  • Speaker:
    1) Mickael Binois (researcher in Team Acumes, Inria Sophia Antipolis)
    2) Areski Cousin (professor at IRMA Université de Strasbourg, and associate researcher in Team Calisto)
  • Abstract:
     – 1st presentation:

    We present an overview of the use of Gaussian processes as surrogate models for expensive-to-evaluate black-boxes. In particular, we focus on calibration (inverse problem) and uncertainty quantification, illustrated by several application examples from various domains.
    – 2nd presentation:
    We explore the abilities of two machine learning approaches for no-arbitrage interpolation of European vanilla option prices, which jointly yield the corresponding local volatility surface: a finite dimensional Gaussian process (GP) regression approach under no-arbitrage constraints based on prices, and a neural net (NN) approach with penalization of arbitrages based on implied volatilities. This presentation will focus on the GP approach. We demonstrate the performance of these approaches relative to the SSVI industry standard. The GP approach is proven arbitrage-free, whereas arbitrages are only penalized under the SSVI and NN approaches. The GP approach obtains the best out-of-sample calibration error and provides uncertainty quantification, which is useful for the assessment of model risk. The NN approach yields a smoother local volatility and a better backtesting performance, as its training criterion incorporates a local volatility regularization term.
    Link to the presentation: PDF 1, PDF 2

8th seminar of Team CaliSto “Modeling and numerical simulations of fluid flows with complex rheology”

  • Date & Time: Thursday, June 3rd, 2021 at 10:00 (CET)
  • Speaker: Rudy Valette (professor in Team CFL at CEMEF Mines-ParisTech, Sophia Antipolis, France)
  • Topic:
    Modeling and numerical simulations of fluid flows with complex rheology
  • Abstract:
    Heterogeneous fluids (pastes, granular, suspensions, emulsions…) generally present a strongly non-Newtonian rheology. They flow very differently from fluids such as air, water or honey. We will show various examples of flows (stretching or compression of filaments, jets, collapses, extrusion…) showing instabilities related to their rheology and discuss with the help of experiments, numerical simulations and simple theoretical analysis the mechanisms involved in these flows. Several examples will be proposed and the interest of numerical simulation will be underlined, both for the understanding of simple phenomena, but also for the prediction of more complex flows, such as those encountered in industry or in nature.
  • Link to the presentation: TBA

7th seminar of Team CaliSto “Sensory navigation using turbulent odor cues”

  • Date & Time: Thursday, May 6th, 2021 at 10:00 (CET)
  • Speaker: Agnese Seminara (CNRS researcher at Institut de Physique de Nice, France)
  • Topic:
    Sensory navigation using turbulent odor cues
  • Abstract:
    Organisms use a variety of sensory cues to navigate to targets like food or mates, and to escape predators. Chemical and mechanical signals are carried long distance by fluid flows, but at the same time they are often distorted and broken in intermittent patches by turbulence. Here we draw inspiration from olfactory navigation in octopuses and mice and ask how can organisms predict the location of a target, using its turbulent odor cues. I’ll show how tracking odor cues for few seconds is sufficient to make predictions, if care is taken into quantifying both intensity and timing of odor detections. I’ll map our predictions on our preliminary data of odor representation in the mouse brain.
  • Link to the presentation: TBA

6th seminar of Team CaliSto “Simulation of the formation of soot particles: toward a better characterisation of these particles.”

  • Date & Time: Thursday, April 8th, 2021 at 10:00 (CET)
  • Speakers: Jérôme Yon (researcher at Dept. Optic & Laser, CORIA, Rouen, France)
    et José Moran (PhD student at Dept. Optic & Laser, CORIA, Rouen, France)
  • Topic:
    Simulation of the formation of soot particles: toward a better characterisation of these particles.
  • Abstract:
    Soot particles are formed due to the incomplete combustion of flames. These particles are not spherical and display a range of fractal morphologies that have to be take into account in various situations (metrology, transport, deposition, …). Tools based on laser metrology are being developed at CORIA to characterize such particles with in-situ measurements. A precise characterisation requires to accout for the fractal morphology of aggregates. To that extent, a code has been developed (called MCAC, for Monte Carlo Aggregation Code) to compute realistic soot aggregates. In this seminar, we will present the issues related to the characterisations of such particles using optical techniques as well as the principle of MCAC. Recent results obtained with this code will also be presented.
  • Link to the presentation: PDF

5th seminar of Team CaliSto “Particle resuspension from monolayered deposits: evidence of collision-induced phenomenon”

  • Date & Time: Thursday, March 11th, 2021 at 10:00 (CET)
  • Speakers: Grégory Lécrivain (researcher at the Institute of Fluid Dynamics, HZDR, Germany)
    and Christophe Henry (researcher at Team CaliSto, Inria Sophia-Antipolis, France)
  • Topic:
    Particle resuspension from monolayered deposits: evidence of collision-induced phenomenon
  • Abstract:
    We present here new experimental and numerical results for the resuspension of microscopic glass particles from a monolayer bed into a turbulent gas flow. With an intermediate surface coverage, here about 10 % of the field of view, we report two distinct detachment mechanisms: collision-induced resuspension at low flow velocities and individual particle resuspension at higher fluid velocities.
  • Link to the presentation: TBA

4th seminar of Team CaliSto “Corrola from drop impacts on viscous fluids”

  • Date & Time: Thursday, February 11th, 2021 at 10:00 (CET)
  • Speaker: Florence Marcotte (researcher at Team Castor, Inria Sophia-Antipolis, France)
  • Topic:
    Corrola from drop impacts on viscous fluids
  • Abstract:
    L’étude des substrats de viscosité variable est motivée par la volonté de réconcilier d’abondantes observations d’impacts de gouttes, qui dans la littérature mettent en évidence des dynamiques de splash très diverses, d’une part pour les impacts sur substrat solide, d’autre part pour les impacts sur substrat liquide. Faire varier la viscosité du substrat sur plusieurs ordres de grandeur permet de capturer la transition d’un régime de splash-sur-liquide à celui de splash-sur-solide, et de clarifier le rôle du substrat dans la formation de la corolle d’impact.
  • Link to the presentation: PDF

3rd seminar of Team CaliSto “Uncertainty quantification in CFD simulations of multiphase flow”

  • Date & Time: Thursday, January 14th, 2021 at 10:00 (CET)
  • Speaker: Aurore Dupré (post-doc at Team Calisto, Inria Sophia-Antipolis, France)
  • Topic:
    Uncertainty quantification in CFD simulations of multiphase flow
  • Abstract:
    In this work, we present recent advances that have been made within the framework of the VIMMP project (Virtual Materials Market Place). We start by introducing the workflow studied (including both the physical aspects, the modelling approaches retained and the numerical tools used). Sensitivity analysis tools are then presented and used to identify the key parameters affecting the observed results.
  • Link to the presentation: PDF

2nd seminar of Team CaliSto “Developments in Feel++ for micro-swimming simulations”

  • Date & Time: Thursday, December 10th, 2020 at 10:00 (CET)
  • Speaker: Luca Berti (PhD at IRMA, Strasbourg, France)
  • Topic:
    Developments in Feel++ for micro-swimming simulations
  • Abstract:
    In this talk, we will focus on the additions and developments motivated by micro-swimming simulations in the Feel++ library. We will talk about articulated swimmers, remeshing and the interfacing of fluid and elasticity equations for a moving body.
  • Link to the presentation: PDF

1st seminar of Team CaliSto “Physical-based probabilistic models for the instantaneous turbulent kinetic energy:modelling, calibration and validation”

  • Date & Time: Thursday, November 5th, 2020 at 10:00 (CET)
  • Speaker: Kerlyns Martinez-Rodriguez (post-doc at Team Calisto, Inria Sophia-Antipolis, France)
  • Topic:
    Physical-based probabilistic models for the instantaneous turbulent kinetic energy:modelling, calibration and validation
  • Abstract:
    In this talk, we present the construction of a probabilistic model for the instantaneous turbulent kinetic energy filtered at a given point in space. For this, we start from the Lagrangian description of fluid flow and derive a mean-field dynamic featuring the square norm of the turbulent velocity of a generic fluid particle. By approximating at equilibrium the nonlinear terms, we recover the CIR-form previously suggested in the literature for modelling the wind speed. In a second part, we establish a calibration procedure for the parameters by employing classical explicit methods and Bayesian inference. In particular, we show the consistency of the estimators and validate the model through the quantification of uncertainty.
  • Link to the presentation: PDF

Comments are closed.